Maurice Quinlivan (Limerick City, Sinn Fein)
Link to this: Individually | In context | Oireachtas source

Members participating remotely are required to do so from within the Leinster House complex only. So far, apologies have been received from Senator Garret Ahearn.

Today we will look at the impact of artificial intelligence on business. In October 2023, the committee reported on artificial intelligence in the workplace. One of the key observations of the committee was that further discussions would be needed to explore the wide-ranging impacts AI might have. The issue of how Ireland can best position itself so as to be ready to face the challenges posed by AI and seize the opportunities AI presents is likely to be a major priority for all stakeholders in this area.

The committee is therefore pleased that we have the opportunity to consider these matters further with representatives from Accenture, PwC and EY. I am pleased to welcome from Accenture in Ireland Ms Hilary O'Meara, country managing director, and Mr. Denis Hannigan, managing director, data and AI; from EY, Mr. Eoin O'Reilly, head of AI, and Dr. Loretta O'Sullivan, chief economist; and from PwC Ireland, Mr. David Lee, partner and CTO, Mr. Martin Duffy, head of GenAI, and Ms Aisling Curtis, market leader, strategic alliances.

Before we start, as we always do, I wish to explain some limitations to parliamentary privilege and the practices of the Houses with regard to references witnesses may make to other persons in their evidence. The evidence of witnesses physically present or who give evidence from within the parliamentary precincts is protected pursuant to both the Constitution and statute by absolute privilege. Witnesses are again reminded of the long-standing parliamentary practice that they should not criticise or make charges against any person or entity by name or in such a way as to make him, her or it identifiable or otherwise engage in speech that might be regarded as damaging to the good name of the person or entity. Therefore, if their statements are potentially defamatory in relation to an identifiable person or entity, they will be directed by me to discontinue their remarks. It is imperative they comply with any such direction I may have to make.

The opening statements have been circulated to all the members. To commence consideration of our matter today, I invite Ms O'Meara to make opening remarks on behalf of Accenture.

Ms Hilary O'Meara:

I thank the committee for inviting me to be here today. I am the country managing director of Accenture in Ireland, and with me is Denis Hannigan, our data and AI lead.

For those who do not know Accenture well, it is a global technology organisation employing 742,000 people and more than 6,000 people in Ireland. We have been in Ireland for more than 54 years. Our global research and development centre, namely, The Dock, is home to one of Accenture's 24 GenAI studios globally and is also based in Ireland.

We work with clients in almost every industry. Most of our clients need to transform and continually adapt their businesses to stay ahead, and we support them to do that. In essence, we help our clients navigate and implement complex change.

Today, we will focus on generative AI, GenAI, the latest form of artificial intelligence. GenAI is a groundbreaking technology, surpassing all previous digital technologies. Our research shows that adopting responsible, people-centric GenAI approaches could unlock $10.3 trillion in global economic value when combined with classical AI forms, such as predictive AI. Despite the hype, GenAI is a game-changer that will revolutionise work and life. It is advancing rapidly. Its potential impact spans all processes and roles, from CEO to front-line workers. Business leaders must understand this new technology and the impact it will have across their industry, and we recommend they position GenAI in their top three strategic priorities. GenAI will transform and disrupt every industry, offering both opportunity and risk. It is important therefore that business leaders navigate decisions with an inclusive, ambitious and responsible strategy.

We have distilled these considerations into three imperatives. First, we must lead with value. Leaders must think holistically, considering how AI impacts end-to-end processes to drive growth, optimise costs and enhance customer engagement. Our advice is to be value led. One should not embrace GenAI for the sake of it, but should be laser-focused on the return on investment for their business and have a clear strategy for how they will harness its power to drive tangible business outcomes.

The second imperative is having the right technical foundations for GenAI, something we in Accenture call the "digital core". Quality data and cloud access are essential. Irish organisations have been slower to adopt cloud compared with peers in other countries. Accenture published research in 2022 which outlined that just 16% of Irish companies believed they had the systems, architectures and structures in place to enable and support innovation end to end, which is relative to the global average of 25%. We advise clients to define their north star and start their journey at a pace aligned with their investment ability and risk appetite. There is no one-size-fits-all approach.

The third imperative is to bring the workforce on the journey. Leaders must focus not only on technology but also on preparing the workforce for GenAI. Getting the workforce ready is of utmost importance and CEOs to whom we speak are particularly interested in this area. AI will have a significant impact on the workplace and employees are concerned about its implications. They worry about job disruption, stress, potential burnout and the quality of technology.

Leaders must focus on two key areas concerning employees. First, it must build trust through regular and transparent communication about the company's plans for GenAI and its impact on employees. Trust in both the tools and the organisation's intentions is crucial. Second, leaders need to prioritise reskilling, starting with themselves. Only one in three leaders feel they have the necessary technology expertise for GenAI, and only 5% of organisations invest in reskilling. All leaders, not just technology and data practitioners, must understand how GenAI will transform their business and then quickly implement learning and development programmes for employees.

We are optimistic about the impact of GenAI on the workforce for several reasons. We believe this technology will drive productivity and make work more enjoyable, leading to greater job satisfaction. Our research shows that GenAI positively affects entry level and lower-skilled workers, boosting their contribution early in their careers. Additionally, the value of GenAI relies on the application of new skills and capabilities, creating new jobs and professions and opening up new opportunities.

While leaders need to work on delivering value, getting their digital core in place and preparing their workforce, all of this needs to be underscored by responsible AI to ensure important considerations such as data privacy, transparency and principles of fairness remain front and central. It is important to be cautious about the widespread use of these technologies without proper oversight. Responsible AI practices are necessary due to the unique considerations of GenAI. While 90% of organisations have AI principles, only 2% have fully operationalised responsible AI. Closing this gap is crucial. Establishing governance frameworks, testing processes and monitoring tools are essential to harness AI benefits while mitigating potential harms. The EU is at the forefront of pioneering efforts with initiatives such as the EU AI Act and the European AI strategy. These frameworks set global examples. Indeed, Accenture references the Act to shape our internal AI compliance programme. Prioritising safe and responsible AI development and deployment is critical.

At a national level, we face a choice between seizing unprecedented opportunities and addressing the risks of these technologies. We have the potential to leverage our national convening power, including policymakers, academia and the business and technology ecosystem, to position Ireland as a leader in the AI era for a successful and prosperous future. Actions that can help amplify this positioning include promoting responsible AI innovation that would position Ireland as a leader in innovating with AI in an ethical manner by ensuring effective implementation of all regulations and broadening our research and innovation capabilities. Developing an AI-skilled workforce would start with aligning our education programmes to prepare our emerging workforce for AI-driven industries and could include incentives or funding for organisations to support continuous learning and reskilling for current professionals, with a focus on the SME sector in particular. Strengthening AI research and industry partnerships would involve boosting collaboration between universities, technology companies and public institutions to drive AI innovation. This could be through funding joint research projects or mentorship programmes to further drive innovation. Co-creating the future would involve establishing a think tank to look at all the ways that Ireland can maximise GenAI. Ireland should lead from the front, encouraging and driving momentum in AI adoption across all parts of society. A failure or reluctance to engage could pose risks to our future economic sustainability.

We believe this is a pivotal moment in which AI is poised to transform and reshape many aspects of how businesses operate. Ultimately, we are embarking on a period of ongoing change and transformation that requires a mindset of continuous improvement and reinvention and which presents Ireland with a new range of opportunities to position for future growth and economic prosperity. We know this is not straightforward, but given the pace of change and the opportunity it represents, we need to act and act now.

Maurice Quinlivan (Limerick City, Sinn Fein)
Link to this: Individually | In context | Oireachtas source

I thank Ms O'Meara. I now invite Mr. O'Reilly to make his opening remarks on behalf of EY.

Mr. Eoin O'Reilly:

I thank the Chair and committee members for the opportunity to come before it today. I am a partner in EY and I am head of analytics and AI practice. I am joined today by EY Ireland’s chief economist, Dr. Loretta O’Sullivan, who will speak soon.

We welcome the opportunity to address this committee on this critically important issue. In EY, we believe artificial intelligence can have a transformative, positive and far-reaching impact on all businesses, from SMEs to indigenous and multinational companies, as well as the people working in these organisations and indeed on wider society. For those who do not know about EY, it employs more than 5,000 people on the island of Ireland, with offices in Dublin, Belfast, Cork, Galway, Limerick and Waterford supporting organisations across a wide and diverse range of professional services from assurance to corporate finance, tax, law, consulting and analytics and AI.

Every day, we work with a broad range of organisations spanning every sector of Ireland’s economy from start-ups and private businesses to well-established Irish plcs and multinational foreign direct investment, FDI, organisations. Our entrepreneur of the year programme has been running for more than 25 years, promoting entrepreneurship and empowering entrepreneurs to grow their businesses on the island of Ireland and around the world. It is this continuous interaction with our varied clients here and across our global network that gives us a unique perspective on how businesses are embracing transformative technologies such as AI from the C-suite to the production floor.

EY has embedded AI into our global technology solutions for more than a decade. Globally, we have invested more than €1.3 billion in our artificial intelligence offering, encompassing services and capabilities for our clients and our people. Here in EY Ireland, we employ more than 300 people in our AI and analytics practice, which brings together experts across artificial intelligence, data analytics and business transformation. These services are to help organisations harness the potential of AI and reimagine the way they do business.

Last year, we launched EYQ, EY’s own private large language model, which operates in a similar manner to other conversational AI models such as ChatGPT or Google’s Bard, allowing our people to ideate, research and create with generative AI. We see EYQ as a co-pilot for our people, augmenting human potential by supporting ideation and research, boosting productivity, automating repetitive tasks and more.

Artificial intelligence is not a new technology, but it is increasingly coming into focus with developments around generative AI, GenAI, in recent years, as other speakers have mentioned. AI is around us every day, from interpreting human speech like Alexa in our kitchens, learning and inference based on a history of events like what we have bought and is recommended to us on Amazon or the generation of documents and information like we are seeing with ChatGPT and other virtual assistants.

From a business perspective and from our interactions with clients, there is real excitement about what AI, and GenAI in particular, can bring to their businesses, what outcomes it can drive and what it might do over the coming years. This is a technology discussion that has vaulted from the server room to the boardroom. In our survey of 1,200 CEOs from around the world, we saw that 99% plan to invest in generative AI and 70% seek to act quickly to avoid being edged out by competitors. Organisations see GenAI as a value creator, boosting productivity and efficiency but also supporting research and new product development, automating repetitive tasks and reducing silos across an organisation and, ultimately, allowing them to better serve their customers. We are seeing engagement across all business functions, from customer service and sales, manufacturing and marketing to core business services such as operations and finance. While the impact may be different depending on the type of business or the area it is focused on, the opportunity is there for almost all.

Second, investment in AI is already happening. Many of the organisations we work with here are already investing heavily in AI or assessing their current maturity to identify where they are going to start. Globally and locally, EY is working across a broad range of areas to support clients on their AI journey. This includes healthcare and life sciences, working with hospitals and on the manufacturing floor to improve yields and products. In logistics, AI is being used to optimise supply chains, reduce delivery times and ensure better efficiency on delivery windows. In finance, banks and other multinationals are making significant investments in AI to train and support the detection of suspicious activity and use AI to automate the review and analysis of complex credit agreements.

AI talent is a top priority for businesses seeking to take advantage of the opportunities of this game changing technology. Access to AI talent and the training and upskilling of existing staff so that they can harness AI in their work is key. That is why it is so important to have skills training and education for people at all stages of their career and in almost all sectors to upskill in AI.

The main challenges our clients are facing are twofold. The first is around data. The quality of AI outputs for a given organisation are proportional to the quality of the data and the digital core that underpins them. That is why getting AI ready and investing early in capturing and organising data accurately is a critical first step that most organisations are currently focused on.

Second, businesses are concerned about how they can adopt AI responsibly and safely. They want to know how they resolve concerns in areas like data privacy, intellectual property, unwitting hallucinations by AI and the essential issue of AI ethics and regulation. Many organisations will be racing to use AI in 2024, but responsible and future-proofed adoption is crucial to avoid pitfalls ahead of new regulation. In this context, the EU’s recently passed AI Act is very welcome, providing guardrails and guidance for organisations and a probably a first-mover advantage to the block.

In summary, technological revolutions are not new, but the economic impact of AI could surface more rapidly than previous technological revolutions, as the implementation and pace of adoption is faster and the scale of the opportunity is very large. This is a very exciting time for the Irish economy. We have some unique opportunities that we should consider but we need to be focused. My colleague, Dr. Loretta O’Sullivan ,will now address some of the macroeconomic opportunities for businesses and the wider economy offered by AI.

Dr. Loretta O'Sullivan:

I thank the committee for having us here today. I want to talk a little about the economy. I will begin by saying that today’s economy is doing well. Ireland is expected to put in a relatively good growth performance this year as inflation tracks lower and the interest rate cycle turns. The labour market remains strong. We have full employment, the unemployment rate is low and the public finances are healthy. That we are in this position after a series of significant external shocks, such as Brexit, Covid-19, war in Ukraine and conflict in the Middle East, speaks to the resilience and can-do attitude of Irish households and businesses. It also reflects the underlying strengths of the economy, including a dynamic FDI sector and talented indigenous entrepreneurs, a skilled workforce and world-class universities and an enterprise-friendly environment.

There is a lot to be positive about but there are headwinds and if Ireland is to keep pace with developments in global markets, we also need to cultivate the seeds of tomorrow’s growth. This means doubling down on efforts to boost the labour force, bolster the capital stock and enhance productivity. We can do this by upskilling employees, investing in necessary infrastructure, accelerating the green transition and harnessing the power of new technologies such as AI.

Along with sustainability, digitalisation is reshaping all of our economies and AI stands at the forefront of this transformative wave. Although the full extent of its economic and social consequences is not fully understood yet, it clearly has the potential to enhance productivity. When it comes to the labour market, technological innovation typically impacts through job creation, where new roles and opportunities emerge; job displacement, where functions become obsolete due to automation; and job transformation, where the nature of tasks is augmented. As AI adoption increases and skill requirements change into the future, it will be important to ensure that enterprise and education policies are in place to promote and support inclusive training and life-long learning.

As the National Competitiveness and Productivity Council highlights, the AI landscape is rapidly evolving and early adaptors stand to gain significant economic, strategic and competitive advantages. EY’s forthcoming FDI attractiveness survey finds that AI would help some two in five international investors expand their mandate in Ireland over the coming years, reinforcing the fact that as a small open economy, this is a must do if we are not to be left behind. Ireland is well positioned in this regard, given that we have a national AI strategy, here for good, in place since 2021, with the most recent update showing progress against all key strands. However, the focus at all levels must be maintained. In EY, we believe that now is the time to invest in the drivers of next generation economic growth, including AI. From past experience, we know that such investment pays dividends many times over, helping to bring about a more prosperous society for all.

Mr. David Lee:

PwC would like to thank the joint committee for this opportunity to contribute to its deliberations on the impact of AI on businesses. By way of introduction, I am a partner in PwC Ireland and have over 30 years technology consulting experience gained in the Irish market and internationally. I am the chief technology officer in PwC Ireland and I am responsible for the operation of our GenAI business centre that we established in collaboration with Microsoft in November 2023. The purpose of our centre is to provide organisations with the opportunity to understand the potential impact of GenAI and AI more broadly on their businesses from an opportunity and risk perspective. We also look to help organisations to assess the ethical, data privacy and human impacts arising from the adoption of AI within their organisations.

Since our launch, we have had over 70 Irish based organisations engage in workshops and discussions in our centre. These organisations represent all facets of the Irish economy, from large FDI organisations through to indigenous Irish companies. They also represent a diverse range of business sectors, from medical device manufacturers to financial services organisations. My observations are based on the aggregated experience of these organisations and a number of surveys that we have conducted on the topic in the Irish marketplace and globally during the last six months.

Before setting out these observations, I thought it would be useful to first briefly reflect on the term "artificial intelligence" or AI.

“Artificial intelligence” is an umbrella term that covers a wide range of adaptive technologies, including machine learning, digital assistants, predictive models and generative AI. Many of these techniques have been around for decades and are well understood and widely used by many of us in our day-to-day lives; just think of the Netflix movie recommender or Siri. However, tools such as generative AI, or GenAI, are much newer, having been in the public domain for less than two years. GenAI offers the potential to make AI capabilities available to a much wider audience of business users than has been the case to date with other AI technologies. While GenAI technology is still evolving at pace and its impacts on businesses are not yet fully understood, the PwC 2023 GenAI business leaders survey found that 76% of respondents expect GenAI to significantly impact their enterprises over the next five years.

The potential impact on businesses of GenAI and other AI technologies also needs to be viewed in the wider context of a series of factors driving significant disruption to businesses. Considerations such as climate transition, geopolitical uncertainty, the macroeconomic environment and technological disruption, including but not limited to AI, contribute to a position where nearly three in ten Irish CEOs who responded to our 2024 Irish CEO survey are concerned about their long-term business viability, a concern that is even more pronounced at a global level, where 45% of respondents expressed a similar concern. The same survey found that the majority of Irish CEOs see GenAI as a catalyst for business change. Nearly two thirds believe that GenAI will significantly change the way their company creates and delivers value in the next three years. Consistent with the findings of the committee’s report on artificial intelligence in the workplace, which highlighted the importance of upskilling and digital dexterity in transforming the employment landscape, the majority of CEOs recognise that it will require workforce upskilling in the next three years.

However, Irish CEOs also recognise that the adoption of GenAI technologies is not without risk. Over three quarters expect GenAI to increase cybersecurity risks in the year ahead. The majority of Irish CEOs expect that GenAI will increase the spread of misinformation, as well as legal liabilities and reputational risks. We are seeing significant interest from businesses in understanding the implications of the EU AI Act and what they will need to do in order to be compliant with its terms. These cybersecurity-related concerns are consistent with the findings of our 2024 digital trust survey, which found that more than half of Irish respondents expected significant GenAI-enabled cyber events in the next 12 months, but also that a majority had plans to deploy GenAI capabilities as part of their own cybersecurity defence arrangements.

While the findings of the surveys referenced above are consistent with our experience in our GenAI centre, it is also important to highlight that the impact across businesses is likely to vary by sector. Ongoing research by PwC, which looks to assess the potential impact of Gen AI on various industry sectors, categorises industries such as software development and pharmaceuticals, which have a material presence in Ireland, as among the high-potential impact industries.

In summary, we are still in the early stages of understanding the business opportunities and the associated risks presented by the application of GenAI technologies. In this context, it will be important to create a business environment, both from a regulatory and enterprise support perspective, that ensures its application to business is done in a manner that is responsible and human-centric while also ensuring the economic opportunity presented is maximised.

Maurice Quinlivan (Limerick City, Sinn Fein)
Link to this: Individually | In context | Oireachtas source

Thank you. I now invite members to discuss the issues with the representatives present. I call Deputy O'Reilly.

Louise O'Reilly (Dublin Fingal, Sinn Fein)
Link to this: Individually | In context | Oireachtas source

I thank our guests for the information they have given us. This is one of a series of engagements we are having on this subject. I am sure I will not be the only one to say it this morning but the more we know, the more we realise what we do not know. That does not mean we are not interested. It is an issue that the committee is taking a particular interest in, which is very welcome, and we do have concerns.

The witnesses said they have a belief that generative AI will make work more enjoyable and lead to greater job satisfaction. How do they think this will be the case? The technological revolution has not necessarily been good for workers. Without the right to disconnect, people are effectively available all of the time. There are workers all around this city, and in every other city and town in every county in Ireland, who are on their bikes at the moment, being driven by algorithms. To have responsible and human-centric AI is important but I am genuinely curious to know how the two can be married, on the one hand, drive, competitiveness, productivity and so on, and, on the other, making work more enjoyable and protecting workers’ rights. That is a tough balancing act.

I will pass over to the witnesses and they can explain how it will be done and we can all take notes. To be serious, it is an issue of real concern. People are having their entire work day - which, when they are on a low income, can take up a considerable chunk of their day - driven by algorithms without any human element. That does not take account of a person's personal circumstances in the way that a human can. The witnesses have expressed a belief that it will make work more enjoyable and I am expressing a degree of scepticism. Perhaps they can explain and elaborate.

Mr. David Lee:

I am happy to take that question. Each of the speakers spoke about the importance of the opportunity being taken advantage of in a responsible way. I can only speak to the experience we have had of the 70 organisations that come through our own business centre. As owners and management within the organisations, they are all acutely aware of the trust equation, and that is trust in terms of the engagement with their customers, but as important is the engagement with their own staff. Deputy O’Reilly is correct in terms of the perception and the coverage. There is a level of apprehension among employees, regardless of what level they operate at within the organisation and whether they come from a professional background or otherwise. What we are seeing increasingly from those organisations is that when they are looking to the governance arrangements they are putting in place to oversee the introduction of the technologies, they are not just looking at that as a technology question and they are seeing it as a trust journey with their employees. They are ensuring that HR is appropriately represented in those governance forums, so the proposed introduction of the technologies is looked at not just through the eyes of the economic equation but also through the impact on employees.

Ms Hilary O'Meara:

I will not repeat what Mr. Lee has said but to build on that, it is very important that you bring your employees on the journey. As I mentioned, that is one of the three imperatives. There are two things. Building on what Mr. Lee said about trust, you have to communicate with employees and let them know how this technology is going to be used, what the impact is on them and why it is important to use it, and bring them on that journey with you to build that trust. They also need to be skilled and re-skilled to be able to use that technology.

I will ask Mr. Hannigan to share a real-life example to give the committee the real meat of how these technologies will help. As a way of thinking about it, every single person in Accenture has been trained on GenAI, from the person who answers the phones through to the CEO and globally. Our leadership team have been immersed for the last six months in deep learning around GenAI so we are confident to talk about it, we understand it and we are not afraid of it. We then bring our people on the journey with us. For every new piece of GenAI, we have a portal that outlines what is going into our workforce around artificial intelligence and what is coming, and we are having that dialogue with employees. I think everybody needs to be doing the same.

From the point of view of job satisfaction, we know we have a lot of job vacancies in Ireland and we know many people have to work incredibly hard to keep up. This technology can help as a service support, a co-pilot or an assistant to people as they do their work. It can save them from doing some of the more challenging and administrative parts of the job and free them up to do some of the more interesting parts.

Mr. Hannigan might be able to share one of the examples.

Mr. Denis Hannigan:

I will try to make it more real and concrete in terms of some of work we are doing with a global beverage company.

In that instance, it has asked us to help it to look at its sales activity and how its sales agents can be more productive and effective in the market. The existing work those agents have been doing traditionally has been intensive in terms of looking at spreadsheets, gathering data and trying to do some analysis on what they need to do on a daily basis. Recently, we have evolved that to having a GenAI-enabled sales coach that agents can use daily. Not only does it give them insights around some of the opportunities that are there and puts some structure on their day, but even more importantly, it helps them interact with their target clients for that day. It gives them draft scripts and enables them to rehearse their pitch by recording their voices onto an iPhone and then getting some feedback on that. It is shifting the type of work they are doing and, ultimately, it will take out some of the administration, research and heavy lifting that is required. It is an example of how the nature of work can change that is also leading to some new roles and activity in that regard.

Mr. Eoin O'Reilly:

I echo what other witnesses have said and I welcome the committee's report that was published before Christmas on some of the things with which we are concerned, such as workplace surveillance and all of those things. There is some really good work done in this regard.

Like Mr. Lee from PwC mentioned, we are seeing many of our clients think about AI and GenAI as an assistant and in the context of augmentation rather than replacement. Many of our customers say they still know that the differentiation in their businesses comes from the people. Their people interact with customers and they need to retain that. While they can have GenAI help, their people are probably still the most important part of their businesses.

Louise O'Reilly (Dublin Fingal, Sinn Fein)
Link to this: Individually | In context | Oireachtas source

The witnesses referred to jobs that are skilled or semi-skilled. That is grand. However, for people who have an app on their phone, which is all that guides their day, the witnesses will understand the fear of that app being somehow empowered to hire and fire without any human interaction. While I respect what the witnesses are saying and understand where they are coming from, let us be frank; there is no one on a bike out there, even today when the weather is nice, who enjoys that work. That work is now being driven by an algorithm. My question is more about how we ensure that the human element is built in. There must be someone looking at the work a person is doing.

In many cases, it will not free people. Perhaps, in an office environment, it will free some people up to do work of higher value and that is very welcome. However, for a lot of people, it will simply remove compassion from the middle layer. This will leave a boss who is far out of reach and a worker - who potentially is an employee or may be classed as self-employed if the rules are being played with fast and loose - being run by an algorithm that has no human involvement. There are concerns in this regard.

It is good to talk about a partnership model but, to be fair, that is in a skilled or semi-skilled office-type environment. When it is taken out and put onto the streets or into a factory, it is very different. I am not hearing that that issue is being tackled. There seems to be an attitude, not necessarily from the witnesses, but more generally, that there are certain types of work and workers who can just be run by this. For me, that is dangerous and devalues and dehumanises the person who is at the very bottom of that chain.

Mr. David Lee:

As for organisations to which we have had exposure, many of those are manufacturing operations; they are not all professional staff. I can only speak to the bona fides of the people who have engaged with us, and they are very conscious. In any environment, there will be rogue operators but I do not think we should design for the minority and not address the opportunity presented to the majority.

When we look at the feedback surveys we get from the folks who go through our centre, one of the most common themes is around governance, not the excitement around the potential of the technology. They are concerned about the reputational risk to their brands and organisations and the risk to that trust relationship with employees, that is, employees regardless of whether they sit in an office or a contact centre or operate on a factory floor. I am not saying it has been cracked and there is an easy answer on it but the awareness, which is always a good starting point, among the organisations we have engaged with is certainly there.

Ms Hilary O'Meara:

To build on the exact case the Deputy was alluding to, technology is already impacting those types of workers and roles.

Louise O'Reilly (Dublin Fingal, Sinn Fein)
Link to this: Individually | In context | Oireachtas source

I know.

Ms Hilary O'Meara:

This is an area in which the Government will have an important role as we go into the future. It is something that we might come back to later. We will need to have the regulation and policy to mind vulnerable workers, as has always been the case. In the context of the advancement of these new technologies like generative AI, that is going to become even more important. That is why it is great to be here talking about this-----

Louise O'Reilly (Dublin Fingal, Sinn Fein)
Link to this: Individually | In context | Oireachtas source

The committee is glad to have the witnesses here.

Ms Hilary O'Meara:

-----for the committee to understand it. It is going to be the Government working with business and academia to figure out this path. Mr. Lee said in his opening statement that it is early days and we need to be vigilant and act when we see things that are going to impact on workers.

Louise O'Reilly (Dublin Fingal, Sinn Fein)
Link to this: Individually | In context | Oireachtas source

I have one other question. I am conscious that my time is tight and I do not think I will have a chance to come back in again because I will have to leave after this. I have to chair a thing at IBEC.

I wish to speak briefly about the importance of high-performance computing in co-ordinating cutting-edge research, developing AI, etc. Do our guests know that Ireland's national supercomputer, Kay, which is hosted by the Irish Centre for High-End Computing in Galway, has been obsolete from November 2023? There is no replacement supercomputer in the pipeline despite the fact that we have known about this for a long time. The State will now have to use the supercomputer in Luxembourg. Such failures of public investment in capital infrastructure for computing and our national supercomputer will have a profound impact not just on the technology sector, but on the sectors the witnesses have spoken about today. I wish to ask, both specifically on this issue and also more generally on infrastructural investments and the deficits that exist, about the impacts that will have on the adoption of AI, how it develops and how it can be built upon, given the State does not have access to a supercomputer, except outside of the State in Luxembourg.

Mr. Eoin O'Reilly:

I was not aware of that. What we are seeing in our clients is that to access these artificial intelligence technologies, development in cloud, which Ms O'Meara mentioned, has been really important. We have seen Irish businesses embrace cloud at a pace that is maybe not as fast as the global pace, but that access to cloud involves a lot of the technologies that Irish businesses need to unlock the value from AI. Those tools are readily accessible through a number of vendors and I am aware that this committee will talk to the technology companies next week in this regard. Irish organisations can access them quickly and start to use them.

Having said that, ongoing investment in infrastructure, not just the technological infrastructure, is critically important in this area and across a number of areas. The Government, with its national AI strategy, is doing some good work with the sandboxes that it is introducing to allow organisations an industry trial. Our more critical ongoing investment will be around the skills arena and making sure that, first, businesses have access to the AI talent and, second, that they also have access to AI-ready workers, if that makes sense. I refer to retraining the entire workforces to be able to use those technologies.

Dr. Loretta O'Sullivan:

From a macro perspective and research that has been done on reaping the productivity gains from all this, it is around preparedness, that is, preparedness in terms of the ecosystem, infrastructure and skilled labour, as Mr. O'Reilly spoke to. Countries that have those ingredients in place and are prepared will benefit more from this than others. Some will be more advanced and will be ahead and Ireland will certainly be in that space. Of course, the productivity gains of this are feeding through to help improve living standards in terms of how economies grow and how we translate that down into improving life for everyone.

Louise O'Reilly (Dublin Fingal, Sinn Fein)
Link to this: Individually | In context | Oireachtas source

I would like to see the evidence linking that to improvements in living standards. Certainly, we will find ourselves in an increasingly difficult place until that supercomputer is replaced. The Chair is telling me that my time is up, so I will leave it at that.

Richard Bruton (Dublin Bay North, Fine Gael)
Link to this: Individually | In context | Oireachtas source

I thank the witnesses for their presentations. Ms O'Meara said that at national level we face a choice between seizing unprecedented opportunities and addressing the risks of these technologies. This is hopefully not so stark as having to choose between the two, although that is certainly where much of my anxiety arises about how we are going to manage this. Another concern regarding what was said is the fact that only 2% of companies have fully operationalised responsible AI. Companies deploying this have not taken on board what looks to be one of the critical ingredients. In addition, only 5% of organisations invest in reskilling. Against a background where we are all saying that this will be an unstoppable technology and that we would like to be leaders in the field, it looks like we are not starting from a good position in the context of our leading companies.

The second issue is to go into what becoming a leader in this technology might mean. It seems that cloud ecosystems, the capacity for computing and so on are at the heart of it. Is there a conflict between Ireland taking leadership here and our ambition to be sustainable? We hear daily that data centres are putting pressure on our energy capacity, but the corollary is that if we do not have data centres, we will not see continuing growth from an Irish base of these transformative companies. How do we resolve that issue?

I am also interested to hear about the disruption. When they talk about sectors that have been disrupted, journalism, law, programming and accounting come at the top. All of the witnesses are in the latter business, so that will presumably be hit. Is it just a question of those people adapting? I suspect that journalism is at the front end and is disintegrating to a degree. As a result of the impact of social media, traditional journalism is almost gone. It has not been a good reconciliation between the power of the technology and what most of us would like to see. I would like to hear the witnesses comments on that.

This brings me back to the issue of how we approach regulation. The EU has done a brave job. Compared with the wild west in the United States and social profiling in China, it is trying to create a framework where we are protected. Will we see the continuing concentration of innovation among a small number of extremely powerful companies, using technology that is so intimate in its influence it will be hard for regulators, other than through principle-based regulation, to keep up and to understand the codes being applied? We have all lived through financial crisis where there was principle-based regulation, so it needs to be more than that. I am interested in that.

The EU is talking about strategic autonomy now, in particular President Macron. As I understand it the EU has the regulation, but it has none of the platforms that are driving at the front end of this technology. Where does that leave Europe? Ought Europe be trying to create something more than just the regulatory model, and if so, how? They are not easy questions.

Ms Hilary O'Meara:

I will start by addressing the first question, which relates to what I said about opportunity and risk. I was trying to emphasise that there is a great deal of fear about these new technologies. Fear can stop you from acting. The Government needs to step up alongside business and other parts of the research-and-development ecosystem to support us to lead in generative AI. If you look at what we have done in technology over the past decade, you can see that we have been at the forefront. We need to sustain that competitive advantage. I have examples of what I would love to see as we go forward. We talk about leading from the front, leadership and taking the lead. If the Government voice was demonstrating that this is something important, it would be significant. What do I mean by that? We talk about the EU AI Act. If we were quick to stand that up and demonstrate that we are policing compliance in the context of regulation and by resourcing the entity that is going to look after our compliance, it would show that we are serious from the point of view of regulation and responsibility.

It is equally about supporting education, and making sure the next generation is literate with and has the skills necessary for the jobs of the future. That will be incredibly important. There are likewise things like reskilling the current workforce. This will be massively disruptive. I am not saying it will be in the next six months, but it is certainly not far away. The pace is much different to previous technologies. Getting on the front foot in the context of looking at how we reskill the population is something businesses will do within their own organisations, but a wider programme like amplifying Skillnet will be incredibly helpful in this area. That is particularly the case for smaller organisations and SMEs for example. We also need to make sure we are at the forefront of policy and regulation. There is much negativity around generative AI, and I understand why. It is good we are having that debate and we understand the risks. Equally, there is a good deal of social value that Irish citizens could benefit from if we get it right. It is about making sure we are not afraid, and that we step up and understand this. I would go so far as to say that there needs to be a Department or agency that is adequately resourced with the right people who obsess about these technologies every day and who understand them. I cannot speak for the others although I am sure it is the same, but Accenture is having to equip itself every day, week and month with these technologies that are moving so fast. Government needs to be in the same position, where it understands them and can set the policy and direction. It can support businesses and employees through what will be a massively disruptive time. That is what I meant about balancing the opportunity and the risk and acting in that regard.

I will speak briefly about responsible AI. Others may then wish to come in on this. Again, it builds on the idea that we need to do this responsibly. I go back to the fact that I am encouraged that all of these challenges with generative AI are being surfaced. We are not going into this blind. We are aware of these risks. Not all of them can be solved overnight. It will take time. Some solutions are starting to emerge. I mentioned that 2% of organisations have principles. That is good. Accenture has seven principles. I am sure PwC has its principles. That is good, but principles do not help anybody if they sit on a shelf. You have to implement them to demonstrate you are taking this seriously. You need to live them day to day. We mean that in organisations you need to be using generative AI so that when designing using generative AI you need to be doing so in a responsible way. You are then making sure that what is built is built in accordance with that responsible design. You are testing in a way that ensures it is responsible. Once you put it in the hands of users, you then have to keep monitoring to make sure it is safe and responsible. You have to govern and put the tools in place to do that. If we do that, it does help.

I will give one example in Accenture. Every client engagement we have, no matter how small or big, has to go through a mandatory risk assessment before we can ever go to do the work with a client. The chair of our responsible AI governance is our global CEO. We are taking this all the way to the top of the organisation. We report to our board globally every quarter on what we are doing with generative AI and demonstrate that we are doing so in a responsible way. It is incredibly important that every organisation is adopting that approach, and that Government is talking about it and making sure that is the approach we are taking. It is about striking a balance between the opportunity and social value we can get from the technology and marrying this to doing it in a responsible way.

Mr. David Lee:

I will pick up on a few of Ms O'Meara's points and maybe link them back to the Deputy's questions. He picked up on the numbers in terms of what appeared to be low levels of adoption of responsible AI, and low levels of reskilling. We need to put into context that this technology was introduced in November 2022. If I categorise what we saw before Christmas with organisations coming in, it was about awareness and education.

Using a golfing parlance, I think 2024 is moving year. This is where people are getting beyond the education stage to actually understanding and piloting the technology in their organisations. I think it is a positive that people are not rushing ahead and are being a little more considered in how they introduce it. To build on some of the conversation with Deputy O’Reilly earlier, it is not a case of getting it out there and seeing how it happens. Rather, it is a more considered view to it because they know they will be with this for the long term.

Looking at the skilling piece, much of the conversation will revolve around what we can do to the current workforce. We need to think further back, to the kids going through first and second level now. We, as well as Ms O’Meara’s organisation, are involved in trying to upskill and introduce people to STEM technologies at an early age so that the opportunity is there for all, regardless of what socioeconomic background somebody comes from. There is a bit of thinking to be done about how to incorporate this into the curriculum so that we get folks entering the workforce being digital literate and not having that reliance on the workforce and employers to take on that obligation. I see with my own daughter coming through college that GenAI technologies are part of what they do. We just need to think about how we get that captured earlier in the cycle so it becomes an opportunity for all and not for, if you like, the minority who happen to have had the privilege of going through an education at university level.

I will pick up on one point and then hand over to Mr. O'Reilly. Sustainability impact was asked about. We did some research that was published last month, and I am happy to share it afterwards. We looked at analysis of the carbon impact of GenAI use. While it undoubtedly is an energy intensive activity, it cannot be looked at in isolation from the problem it is looking to solve. In many cases, the power and capability of GenAI is being applied to address wider climate-related matters. One needs to look at the end-to-end equation rather than just the impact of itper se. I am happy to share that material separately.

Mr. Eoin O'Reilly:

I will pick up on one or two points. On Deputy Bruton’s regulation points, the early stage is a point well made by Mr. Lee. The Act is only starting to surface, so to say that people have not responsibly adopted AI, we are still early in that cycle. They are starting to look at things. The Act asks them to think about their inventory of models, the inventory of AI they are using and to classify them according to risk.

Richard Bruton (Dublin Bay North, Fine Gael)
Link to this: Individually | In context | Oireachtas source

We have seen the technology with journalism, for example. If you do not adopt, it will you shape you and hollow out your business. You will not get the chance.

Mr. Eoin O'Reilly:

Correct. It needs to happen quite quickly. We are seeing it in our own organisation. The Deputy mentioned accounting and lawyers. There is change coming and there is no doubt about that. There is no doubt about how some of those jobs, skills and roles will change and how they will need to adapt. We see that positively. As an organisation that employs many accountants, many of our people are quite excited about some of the tasks – the manual labour-type activities and repetitive tasks – where AI will have a bigger role. It allows them to concentrate on the higher risk areas and the more judgment areas they want to be doing.

On regulation, I welcome that in the Act there is differentiation between companies that are using AI and this idea of the frontier – the big strategic models the Deputy mentioned. The regulation cannot be just light touch. The agencies around Europe will need to get in behind us, look at some of the models and see that the controls the organisations are documenting are actually put in place, particularly in higher risk areas such as healthcare and beyond.

Finally, on sustainability, I agree with Mr. Lee that we need to look at the benefit AI can bring to the climate equation. Optimising use of electricity and water is something we are seeing. It is almost like ancillary services around the data centres are using AI heavily to try to balance and run them efficiency. That could be taken into food production, waste and all those kinds of areas, which are a big part of the sustainability equation.

Richard Bruton (Dublin Bay North, Fine Gael)
Link to this: Individually | In context | Oireachtas source

I have exhausted my time. Those were very interesting answers. I will come back in again.

Marie Sherlock (Labour)
Link to this: Individually | In context | Oireachtas source

I thank all the witnesses. This is an important conversation and we will have to return to it many times again.

Mr. Lee spoke about the cyber risk and the reputational risk. I wish to ask specifically about the management and workers’ rights risk. We have a situation now where the development of tools within the workplace are beyond human surveillance in terms of algorithmic management. We have AI assisting workers in producing output. I wish to ask all the witnesses about the specific safeguards that need to be put in place. It is no longer manager and worker. We have a third party in the room, which is either generated or created in-house or by another organisation. What views have the witnesses’ organisations formed with regard to those specific safeguards? We would take the view there needs to be a legislative safeguard put in place but I would like to hear the witnesses’ views on it.

Mr. David Lee:

At least the EU AI Act is a start in that it is prescriptive in some of the activities that are seen as being at an unacceptable risk level, a number of which touch on the areas the Senator talked about. Regarding the safeguards individual organisations need to put in place, if we followed the methods and approaches we are seeing being adopted, you need to link the purpose for which you will use a technology with the technology. Having a view on the technology in the absence of the proposed purpose is a gap. What we are seeing organisations do and what we encourage them to do is assess the risk on a use-case by use-case basis. The risk is not just a cyber or technical risk but a risk to reputation, trust and workers’ rights, for example. Whether a worker is a white-collar worker, professional staff or so on, they all have the same concerns. I think I can speak to the other organisations here as well. We are not operating in a vacuum from the rest of the world. Our own colleagues have the same levels of apprehension with regard to what GenAI means for our roles. Deputy Bruton correctly said that some of the most impacted areas are our own industries. I do not refer back to the answer I gave to Deputy O’Reilly but I am encouraged by the fact the folks coming into our centres have at the forefront of their minds how to do this in a responsible way. They have worked hard to build their organisations, build trust with their employees and build reputations in the market. They want to ensure they take advantage of this because standing and back not engaging will just very rapidly result in a deterioration of their competitive position, which is good for nobody. They absolutely are doing it in a responsible way, which might mean they are being a little cautious with regard to the first steps and the pace of progression, but I think that is to put the right foundations in place so that when they do engage, it will be in a more thoughtful and sustainable way for the organisations.

Mr. Denis Hannigan:

I acknowledge and reiterate the point that it is a very nascent technology. It is important we are talking about and recognising many of these risks and issues. We are seeing that these require new ways of working and new approaches to mitigate and manage those effectively. To use one client example, if I may - I know it is not quite the scenario alluded to in the question - we are working with an organisation on its chatbot capability and infusing GenAI into that to make it much more engaging for the customers who use that chatbot. That has recently gone live over the course of the past six weeks. It has required a very different way of working for that organisation. This technology is non-deterministic, which means that if you ask it the same question multiple times, you will get different answers on some occasions. You need to rethink how you will test and how you will make sure you are comfortable with the guardrails around that. Our teams had to do a lot of fine-tuning of that particular model, putting in 150 FAQs and putting the right tone and controls in place to ensure this chatbot would respond in a manner that was in keeping with what the organisation was hoping. One of the learnings was that the hardest aspect of the entire project or initiative was bringing some of the stakeholders from privacy, security and risk on board on that journey and ensuring all different stakeholders were comfortable.

That organisation now feels better equipped as it looks to scale up into other areas and the feedback has been positive so far. The key point is that it is a new way of working. It requires a new mindset and new approaches and is something we need to take seriously.

Mr. Eoin O'Reilly:

We all recognise that the technologies now available are much more pervasive, capture more information and have greater capabilities to do things that we do not want people to do. However, we should also recognise that, over a long number of years, data has been available on employees, on how they are moving around the office, on how they are using email and all those types of things. Just because companies have it does not mean they should be doing something with it. We need to take the same approach when the technology is more powerful. As Mr. Lee said, as we are thinking about the potential use cases, we need to think about it in the same light. Just because they have the information and tools does not mean they should do something with it. Organisations need to be very careful in identifying the risks of every use case we define for AI and how it will be used day to day.

Marie Sherlock (Labour)
Link to this: Individually | In context | Oireachtas source

The AI Act is very much about product regulation and much less about the operation within businesses and workplaces. I welcome what the witnesses have said.

Much of the research on AI is about the expected future use. I have not seen any comprehensive assessment or survey of the current use within workplaces. The witnesses are living and breathing this every day. What share of businesses have adopted AI in terms of a whole-of-business use in terms of output? What share of businesses are using it to augment existing tasks? It would be very helpful to get a sense of the picture in Ireland.

Mr. Denis Hannigan:

There is no one common answer, particularly when taking a sectoral view of this.

Marie Sherlock (Labour)
Link to this: Individually | In context | Oireachtas source

Have any of the organisations represented here undertaken a large survey to give us a comprehensive sense of what is happening?

Ms Hilary O'Meara:

In every organisation the employers are using it at the knowledge worker level. Organisations need to have principles and adopt responsible AI. Employees have this and it is successful for everybody. Employers must assume their employees are using it to do their work. It is easily accessible. If any organisation claims employers are not using it, that is not correct. Therefore, it comes back to these points relating to education, transparency, reskilling organisations and then putting the guardrails around the technology. For example, the uptake of Microsoft Copilot has been quite good. We are using it in Accenture. It is really helpful in speeding up some of the administrative tasks. Much of the work in Accenture is now driven by artificial intelligence and generative AI. It has not felt like it has been oppressive; it has been the opposite.

We are in the early days. We are talking about the edge cases of the difficulties of this technology. It can do a considerable amount of good and it presents considerable opportunity and value. The overall message we are trying to relay is that if we do not seize the opportunity and are so afraid, we could lose our competitive advantage. We need to find the balance between being responsible and capitalising on the value these technologies can bring.

Mr. David Lee:

If I were to paraphrase Henry Ford when people were looking for faster horses, he gave them the Model T. In terms of generative AI at the moment, we are probably giving people faster horses. It is using the technology to drive out inefficiencies in day-to-day activities. The next wave will be the more destructive piece where people consider the impact it has on the business model. We will be moving from the horses to the cars. People are learning in a reasonably safe environment. As Ms O'Meara said, in every organisation people are using it. Individuals without any diktat from the management see it as a beneficial aid to their work. Obviously, we need to do that in a much more controlled way. If we did a survey now, it would be very different from what we might see in 12 months' time. As I said earlier, 2024 will definitely be a significant year in going from toe in the water to more material significant deployments.

Mr. Eoin O'Reilly:

There are some great examples beyond the knowledge workers where we are involved day to day. We can think of some of the healthcare applications, such as remote monitoring of patients and detecting falls. That is AI in action. Mr. Lee mentioned the application on the production floor earlier with factory workers being assisted by AI in trying to get better yields and trying to understand the information. These are all really good use cases that are emerging and starting to get to the heart of what businesses are doing and how they create advantage.

Mr. David Lee:

There is a general concern that it might accentuate existing divisions, whether at a gender level or a socioeconomic level. We have a client who is using it to address the significant under-representation of females in its technology community. It is using it to generate job descriptions that would be more attractive to female candidates than to male candidates. That is a very simple but powerful application of it.

Ms Hilary O'Meara:

Coming back to what Mr. Lee said about his clients, every client we work with has been into our generative AI studio not once but multiple times. There is an appetite for this. Some of those clients are figuring out their strategy and plan, where they might start and how fast they might go. Some are starting to learn and experiment. These use cases and test cases are very different from what has gone before. There is so much learning even in doing that. Others are moving to genuinely embedding it in their processes and services, such as the chatbot example Mr. Hannigan gave. There is a huge spectrum of where people are at.

However, we cannot be the fast followers we previously thought we could be in other technologies. This is moving so fast that if we do not get on the bus, we could get left behind. I genuinely mean that. Everybody needs to be embracing, understanding, learning, educating themselves and starting to adopt it. We need to understand the pace at which it will move and the pace at which it will disrupt and get ahead of it. I include government in this. If it is not understood, how can we make the right policy decisions? How can we support reskilling people in the right ways? It is coming very fast. We really need to keep going.

Dr. Loretta O'Sullivan:

In EY we work with a lot of inward investment. Regarding the wider competitiveness in Ireland and our place on the world stage in punching above our weight, we do a survey every year on FDI attractiveness and what makes Ireland attractive. We ask the investors about technologies. Typically, cyber tops that list and AI has been moving up it. It went from 32% in our previous survey to 38% in our current survey which will be published in a couple of weeks. As it is here and it is moving up, we need to be on board with it. As the other witnesses have said, we need to do that in a fair and responsible way. It is about balancing innovation and regulation and taking everybody forward with this.

Matt Shanahan (Waterford, Independent)
Link to this: Individually | In context | Oireachtas source

I welcome our guests this morning it is very timely to have representatives of three of the leading agencies that integrate with government. They will have a pretty important position to use their learning to influence the Government. I am more interested in high-level policy than the individual activities. We could spend all day talking about how AI cuts across all the different things.

My background is in more of a micro business area. I was looking at some of the opportunities and threats. Under opportunities, I have put down transformative for business, increased efficiency, product innovation and the next generation of employment. It is an unstoppable train and we need to be on it. The threats are the job displacement and potentially making humans obsolete. Will future decision-making be done by man or machine? There is business disruption and the whole business model changes. That is the piece we are in. Even though we all say there are great opportunities and great threats, none of us really knows where it is going because ultimately the whole nature of AI is that AI itself becomes the thinking and is creating the shift.

We accept that Ireland needs to be at the vanguard of this because we obviously host some of the world's largest technology companies based here and we want to keep that competitive advantage.

At the same time, we also have to understand it is going to have major cultural, social and economic shifts. How are we going to manage all of this? Part of it is retraining and reskilling. We have to understand there is a significant portion of our workforce who may not be able to adapt to the degree necessary to try to keep themselves in a job, let us say, and that a lot of jobs are going to be displaced. From the point of view of the witnesses, at the top threshold level, what will they be saying to the Government regarding our national capability? This includes our requirements relating to energy supply, which we know; data centres, which we need to be driving on but cannot, it appears, because of the grid; and the regulatory environment, on which at the moment we are going back to Brussels. That may not be enough. We have to be doing a lot ourselves. The key challenges here are around security and the protection of data, data integrity and information integrity. As politicians, that is going to be very significant. Even in the upcoming local elections, we can see already what is happening. In the context of those areas, what will the witnesses be saying to Government in terms of their analysis to guide Government and departmental policy here to make sure we make the right decisions appropriately and at the right time?

Mr. Eoin O'Reilly:

I am happy to kick off on this one. Dr. O’Sullivan may wish to come in as well. I thank the Deputy for the question. I will make some observations in a couple of areas or mention things we are saying in interactions with the Government. First, he will hear loud and clear from all of us here about skills. This is not to fixate on the AI talents, those who do AI; it is also about enabling the rest of the workforce. Some of the programmes that have been launched as part of the national strategy are really welcome but I do not think we can do enough of that in skilling up the workforce to be AI-ready. The Deputy referred to SMEs. We do a lot of work with them. I am blown away by their enthusiasm for some of the tools. Sometimes, they are less risk-averse and they will have a go. I was with a packaging company two weeks ago that was using GenAI to respond to a request for proposal, RFP. One of the employees said it took him ten minutes to prepare a proposal that would normally take an afternoon. There are great simple applications out there. That skills piece is huge and we need a continued strong focus on that.

On the regulation side, we obviously need a strong embrace with the AI Acts. Having trustworthy and responsible AI is really important. It is really important we get that right. There is perhaps a question for us, given the lead role we played in data protection, whether there is something there about which we should be thinking from the perspective of Ireland or if we could play a bigger role in Europe in regulation and compliance of AI.

To move to the multinational sector, we have to recognise that we have done very well in the attraction of multinationals and the types of jobs we have there. There is a protect piece there, quite frankly, regarding the types of work that happen in those organisations such as in finance and manufacturing. We need to make sure we are AI-enabling and protecting those investments. In the context of how those workers execute in their environments, it is really important that AI is to the forefront because there is a lot of competition in those organisations for AI investment and other countries are working very hard. There may be an opportunity, as someone mentioned earlier, around Ireland as a destination. Are there particular places for which we can make a play? From a pure numbers perspective, we should not think we will compete with India or China for pure engineering talent, but maybe there are certain areas where we have been very successful, such as in life sciences or in the financial services sector. There may be something there where we could play a bigger role.

Finally, we should not forget about the other skills relating to AI in which Ireland has a distinctive advantage. I refer to creativity, collaboration and communication. We underestimate how important those are in a technical discipline and we should be really confident about those kinds of capabilities and how we can built on that to attract more investment to Ireland.

Ms Hilary O'Meara:

We have already discussed the detailed points but if I was to take it up a level, reflecting on the Deputy's question, the State has taken quite a good number of actions already which are encouraging. We have the AI strategy and the AI council. Government Departments are mobilising, implementing and experimenting with generative AI. There is an awful lot they are learning and understanding. We have started this journey. However, as we have all mentioned, it is moving so fast that the question is how we bring it all together. Where do organisations such as ours go to give this advice the Deputy is asking for? We need to consider where is the central agency or function that is obsessing about this technology and understands it to a level of detail. Then you can get into great conversations with organisations such as ours and feel a responsibility and an accountability to do something with the information you have given. We need to be acting across all of these things. I would really encourage a consideration of that. We will probably have a new programme for Government at some stage next year and we would love to see these technologies, including generative AI, playing a really strong feature in that. Frankly, it should and needs to feature strongly. There is huge convening power in Ireland, as I mentioned in my opening statement, where we work together and we collaborate. All the tech firms are here in Ireland and we have business, academia and Government. If we all came together, there are some really powerful things we could do. However, we need to have somewhere to go to engage with Government on that, and where that space is situated is the question.

Finally, coming to the list of areas which the Deputy is concerned about, such as national capability and data centres, these are all the questions every government is asking. We do not all have the answers to these questions yet but we should be talking about them, collaborating, taking steps based on our understanding, and also engaging with other governments. We not alone in trying to figure this out. Having the EU there and engaging with it is very powerful. It comes back to that final question of taking the lead in the discussion around how the EU AI Act is going to be implemented. We have the Act but now we have to implement it. If we can step up and demonstrate that we are really taking that seriously, that is a great signal for Ireland.

Mr. David Lee:

I do not want to repeat what my colleagues have said but I will build on one of the examples that Mr. O'Reilly gave. I know the Deputy has a particular interest in the SME sector and there is a great opportunity for the appropriate application of AI to level the playing field. I definitely think there is a role in ensuring the SME sector is given the appropriate support. Our experience has been similar. Some of the most innovative ideas have come from the smaller organisations. It is about giving them that support to be able to harness that technology because it will level the playing field for them.

Matt Shanahan (Waterford, Independent)
Link to this: Individually | In context | Oireachtas source

When the witnesses are engaging with the Government, as I am sure they will from time to time, it is very important that the Government looks at the tax situation and how we can incentivise investment. That is going to be the biggest problem for small businesses to invest in the platforms and technology.

Mr. Eoin O'Reilly:

On that, I have certainly heard the feedback loud and clear from our clients about the changes to research and development incentives. Two or three years ago, the bar was high in respect of the intellectual property creation, almost, for research and development grants. Now, digital and AI investments are starting to be much more accessible, which is very welcome from businesses.

David Stanton (Cork East, Fine Gael)
Link to this: Individually | In context | Oireachtas source

I apologise for being late this morning. I was in the Dáil Chamber. There is some news that may be of interest to the Chair and is, perhaps, related to the issue being discussed here. I was pushing this for quite some time with the Government. It has now decided that spouses and partners of intracompany transfer employment permit holders can work here straight away. They do not have to stay at home. Our witnesses may be interested in that as well. It is operational as of today. I have been pushing it for quite some time and the Minister for Justice, Deputy McEntee, informed me during Topical Issue Debate in the Dáil this morning that it is going to happen. I know we brought it up here on a few occasions as well and we were very supportive of it. It is actually done now and I am thrilled with that. I apologise again for being late.

I thank the witnesses for their presentations. Many questions have been asked already and I will not go over them again. Ms O'Meara spoke about creating a think-tank to look at the ways AI could be maximised. Will she expand on what a think-tank would look like? How will it be done, who would be in it, how would it work and all the rest of it?

Ms Hilary O'Meara:

It is really about reflecting on all the opportunities there are in Ireland.

Ireland has led the way over the past ten years in our economic growth on the back of our investment in technology and other industries. We do not want to lose that competitive advantage. When I pay attention to what is going on in the advancement of technology, there are so many big, open questions the world is grappling with. We are still figuring them out because it is early days in this new generative AI journey. Perhaps it is time for us to convene. We have so many of the tech firms working in Ireland. We have business and we have academia and the Government. As we are a small country, we are quite agile and we are good at engaging with one other. Is there a way to bring that power together and debate some of these issues to bring? It is conversations like this, just probably on a bigger scale that would bring clarity. How it would be done would be determined by whoever would take on the responsibility for doing so. I know there is the global economic summit happening next week at which we are convening business, academia and Ministers. On the global stage we will be debating some of these topics. We could do a very concentrated version of that and come up with some really good ideas. If we have learned anything with our research and development centre, the Dock Innovation Hub, it is that when different people from different backgrounds are brought together to talk to one another, great ideas come. That would be of huge benefit. We could then choose the ten best ideas coming out of the process. I would love to see it being action orientated. There is a huge opportunity for us to do something together. That is my suggestion.

David Stanton (Cork East, Fine Gael)
Link to this: Individually | In context | Oireachtas source

It is a very good suggestion. It is something we need to start seriously looking at. This whole technology area is moving so fast it is being called the fourth industrial revolution. I have been looking at some stuff recently. From the time paper was invented to when the printing press was invented, took 75 generations. This technology has moved forward in less than half a generation. It is moving so fast that it is absolutely scary. It has been described as a tsunami of tech.

I want to ask the witnesses from PwC about red teaming 101. It has to do with testing. It is an attack technique used in cybersecurity to test how an organisation would respond to a genuine cyberattack. Have the witnesses come across this technique? Are many companies in Ireland using it?

Mr. David Lee:

It is not something I am qualified to comment on.

David Stanton (Cork East, Fine Gael)
Link to this: Individually | In context | Oireachtas source

Okay. It is an interesting concept whereby companies get a third party to simulate a cyberattack.

Mr. David Lee:

Yes, sorry, the terminology confused me. Most organisations now perform this as part of their standard business continuity plan annually. They simulate ransomware attacks and the impact that has on business operations. They also simulate the response to this, up to and including deciding whether they would pay a ransom. Unfortunately, it is a far more frequent event than makes the press. We support a number of organisations that have successfully navigated their way through not just a simulation but also, unfortunately, real-life events. Others may want to comment on this but our experience is that it is very much on the board agenda of most organisations to be satisfied that they have a very robust position in being able to prepare for and deal with an incident if it occurs.

David Stanton (Cork East, Fine Gael)
Link to this: Individually | In context | Oireachtas source

We are all concerned about the fact that white-collar workers will be more impacted than blue-collar workers by this technology. I was a teacher in a former life. I have seen that GenAI can actually become a scarily good teacher. It also works well for interpretation skills. We can speak into this and it will speak back in any language required. It is a bit like what they had in Star Trek long ago. This is science fiction becoming science fact and then some. I think it was called a communicator or something.

Matt Shanahan (Waterford, Independent)
Link to this: Individually | In context | Oireachtas source

The universal translator

David Stanton (Cork East, Fine Gael)
Link to this: Individually | In context | Oireachtas source

Well done. Deputy Shanahan has it. I was just testing to see if he was paying attention. We now have a universal translator. It goes on and on. So many jobs which we cannot even guess at now will be impacted by this technology. Would the witnesses agree that the legal profession, among other white-collar professions, is concerned about this?

Mr. Eoin O'Reilly:

There is no doubt that there is general agreement among many of the reports that our professions will be affected. There is a lot of opportunity, if one wants to put a positive spin on it. In any situation where large amounts of information are being processed, there are opportunities. All I can say with certainty as our organisation embraces it, including in its working with law firms, is that it is a very positive embrace. Many of the tasks where productivity could be improved are not top of the list of favourite tasks that people perform. Our people are telling us it is a good thing because it frees them up to spend more time with clients and doing interesting work. That is actually a really good career path.

The Deputy mentioned professionals such as teachers, doctors, nurses. We cannot forget about the empathetic skills which are so important in those roles. AI will have a really big role to play in roles such as teaching assistants, for example. I see it with my own kids doing quizzes with Alexa and how they can learn things from that interaction. We need to think about the possibilities in a hospital environment or a GP's surgery, with the ability to capture information to help with diagnostics and recommendations. It still has to have a human at the centre, however. I do not see a situation where we go to a clinic and see a virtual doctor. That is not what we want. I see the AI as an assistant, with the other human skills being incredibly important as we embrace this journey further.

Ms Hilary O'Meara:

The only thing I would add is that we do not know the future of this. It is to be re-imagined and many new jobs will be created. We do not yet know how teachers will perform their jobs in future. We have to imagine it and that is the journey we will all go on. We need to communicate, be really transparent and talk about this because in a vacuum we will be filled with fear and everybody will be worrying. This is something I constantly talk to my employees about at Accenture. We are constantly putting tools in their hands to get them comfortable with the technology and to understand it. We need to make sure that Ireland does not leave a vacuum that fear replaces and we get a lot of negative narrative. What we should do is embrace and accept this, reskill, keep an eye on it as it moves along, quickly pivot the skills to the new areas and get ready for this massive disruption. Communication and talking about it is really important.

Mr. David Lee:

The impact on work is more nuanced as well. Over the past ten to 15 years we have seen activities move out of Ireland to lower-wage economies. We now see examples of people using the power of GenAI to bring those activities back into the organisation because the labour arbitrage piece is no longer relevant. It is quite a nuanced discussion. When the word "impact" is used, people generally associate it with a negative. In many of the areas that are most impacted, we have skill shortages. If the productivity of a software development team can be improved by 40%, that is good news for everyone. It is not anyone having their job displaced, it just means that more gets done more quickly. When we think about impact, we need to see both sides of the equation.

David Stanton (Cork East, Fine Gael)
Link to this: Individually | In context | Oireachtas source

We could talk about the impact on politics, which probably does not impact on the witnesses. This is being created by businesses in many ways. My good friend here could be portrayed as saying something terrible just before polling day. That would destroy his political career and he would not be able to do anything about it. The videos are so realistic. We need to start educating people to think twice before they accept what they see.

Ms Hilary O'Meara:

Exactly.

David Stanton (Cork East, Fine Gael)
Link to this: Individually | In context | Oireachtas source

This is hugely important. It worries the hell out of me because any of us in politics could be impacted by this and we could be destroyed.

Ms Hilary O'Meara:

Obviously that is a very real challenge at the moment. Yesterday I clicked on an email in Accenture where I had a five-minute video training in which one of our leaders was speaking and then it was not that person; it was artificially generated. It was a quick education on the things to watch out for that demonstrate something is a fake. I had another email two weeks ago from our CEO where it was explained how she had been mimicked by AI and the things we need to be vigilant about.

We are sharing these stories pretty regularly now in Accenture to ensure people realise they cannot believe everything they see and read and they need to check. This goes back to the point about communication and debate at a national level as well. We need to explain this to people so that they understand it. When we do not do that, the vacuum is filled by negativity and fear.

Mr. Eoin O'Reilly:

I would like to make a quick comment on that, if I may. Politics is a great example and journalism is another, where trust is going to become a bigger part of the equation going forward. We need trusted information and trusted sources. We are now seeing a proliferation of information and we have not quite caught up with our regulations. We need trusted sources and curators of information to train the models so that they are trained in the right things. We also need information sources we can trust. I know for certain that for us as a brand that delivers trust and assurance services, we are seeing that our clients want to talk more about this and about how to get trusted information correct in their organisations and get it out into the public domain as well.

David Stanton (Cork East, Fine Gael)
Link to this: Individually | In context | Oireachtas source

I will ask one final question, if I may. We know this can be used to create miracle drugs and do an awful lot in medicine. It can edit DNA and people can even do that now in the back of a garage. They do not need a laboratory. Anyone can do this because it is open source. With respect to rare diseases, it can also create pathogens. That is the other, dark side of this. No one country can actually deal with this. Do we need some kind of global, non-proliferation treaty type agreement to ensure, as best we can, that the real dark side does not emerge?

Mr. Eoin O'Reilly:

That is certainly not something on which I am an expert but the Deputy's point is well made. We are starting to see the forming of the different regulatory approaches. We have seen the US, the EU and the UK approaches develop but we are not there yet in terms of global agreements, although there is a need for them. There are quite different philosophical approaches being taken in some countries, depending on their political dispositions. That is not going to be without its challenges, even though the Deputy's suggestion is a very good one.

Ms Hilary O?Meara:

Collaboration at a national level across countries is really important. Again, it is about dialogue because we are all grappling with the same challenges. The summit that was held in the UK nine months ago brought leaders together and Ireland was represented there. Participants were talking about these challenges and continuing that dialogue will ultimately lead to something of the nature of a global agreement being formed. That is really important.

Maurice Quinlivan (Limerick City, Sinn Fein)
Link to this: Individually | In context | Oireachtas source

I thank our guests for raising the issue of politics and the misinformation that might occur during elections. That is really important. Last year the Mayor of London was undermined in a video that even his own family believed was him speaking about a contentious parade that was about to take place. Police were deployed on the streets to prevent rioting and that all came from a fake AI video. Deputy Bruton is next.

Richard Bruton (Dublin Bay North, Fine Gael)
Link to this: Individually | In context | Oireachtas source

Our guests emphasised several times the challenges for our education system. It strikes me that our emphasis on memory-based assessment is a real obstacle to changing the way we think about using various technologies. What are countries that are trying to prepare children for this sort of transformation doing in terms of curricula and assessment that we might learn from? Our guests also spoke of us taking a leadership position in implementing the new regulatory regime and trying to turn it from principles into action. How well equipped are we to do that? We have had Coimisiún na Meán before us previously. Would it need to be substantially beefed up from where it is now, if that were to be a realistic ambition? Would it have to poach staff from the likes of Accenture and EY to establish that sort of a base?

Mr. Eoin O'Reilly:

On the education system, we are seeing different approaches in various countries and, in fairness, we are seeing the seeds of this in Ireland as well. The most important skills for this discipline are the art of problem solving, curiosity and setting the mind to certain types of tasks. Early efforts around coding languages and programming become really important because they get our children to start thinking about approaches to problem solving, including breaking a problem down into smaller pieces. The other skills we should not forget as we work through from early education to university are the ones I mentioned earlier like communication, collaboration and additional technical skills.

Richard Bruton (Dublin Bay North, Fine Gael)
Link to this: Individually | In context | Oireachtas source

Is it a question of grafting on to what is there or is it a more fundamental shift in terms of getting away from the high-stakes leaving certificate model, with students cramming everything into their heads for three hours?

Mr. Eoin O'Reilly:

I am not sure that prepares our workforce for the AI opportunity.

Richard Bruton (Dublin Bay North, Fine Gael)
Link to this: Individually | In context | Oireachtas source

Is there anyone blazing a trail that we should be looking at in the context of the educational dimensions of this?

Mr. Eoin O'Reilly:

I have seen some interesting things in Singapore and how they have embedded some of those problem-solving-type disciplines at a much earlier stage. I can send on some information on that to the Deputy.

Ms Hilary O'Meara:

I could not talk to education specifically because I am not an expert but we can do some research on it. Every country is figuring this out and there are some new things happening, even in Ireland. The universities here are looking at how they would respond. This is one of the big questions with which we have to grapple. I suspect it will change and will need to change. To go back to things we could do locally, if we were to convene a group that is focused on how the education system is going to be transformed, with people from business contributing to it, that would be incredibly helpful. As I said previously, it is early days and nobody has all of the answers. We can learn from what other countries are doing but then we have to decide what we are going to do here. I would love to see the Government taking action on this and convening a group to work on this because it is an area that is going to change and be disrupted.

Mr. Denis Hannigan:

I would like to comment on the readiness from a regulation perspective. There is an onus within the EU AI Act to set up a new national competent authority, a new sandbox to facilitate some innovation and enable SMEs to determine their own risk approaches. It would be really encouraging to see appropriate investment going into those areas so that Ireland can position itself at the forefront and be a leader. We see some other governments acting. Mr. Geoffrey Hinton is now working for the UK Government and providing advice. It is not about single individuals but about building capability in some of these areas so that the Government is better informed.

Dr. Loretta O'Sullivan:

If we come at this from a macro picture perspective, we see that sustainability and digitalisation, of which AI is part, are reshaping our economies. They are going to be fundamentally different over time. We are at a point now where it is early days. As it emerges that we need new skills, particularly in the context of the sustainability piece and green skills, things will be quite different going forward. Technology will change and we will need new skills in terms of AI and, as they come to fruition and as we discover what those new jobs and new needs are, that will start to filter through. It will start immediately with those who are in the workforce having to adapt, upskill and retrain. At the same time, we will be looking to what is coming and allowing that to trickle down through the education system. It is to be hoped that will happen in tandem so that we are ready when this really takes off. Again, to speak to the point about collaboration, this is related to that old idea about joined-up thinking. It involves different Departments, different areas of government and different areas of society and business coming together to try to forge a way through. This is something we are shaping now and, collectively, we can shape it. The path of least resistance will be the worst outcome. It is hard work and smart policy interventions that will give us the better outcomes.

Mr. David Lee:

One of the roles I have within PwC is that I sponsor our inclusion and diversity from a gender perspective. This is always a challenge from a STEM perspective, where women are significantly under-represented. In some of the research we carried out last year on the perceived likely impact of, or attitude towards, GenAI among women, they are less optimistic than men about the impact of AI on their careers. Currently, among the people who are skilled in AI, our research shows 22% are female. If we look at the cohort of people who are making the decisions on the application of that technology, such as the CIO and CTO population in the US, 8% are female. We need to recognise that in the context of what we are doing from an educational perspective as well such that we will engage early in the system with everybody regardless of gender or socioeconomic background. It would be a disappointing outcome if the adoption of GenAI just accentuated existing problems. That would be an interesting perspective in the context of the think tank Ms O’Meara mentioned to make sure we can do something innovative in that space that uses an opportunity to bridge that gap rather than widen it.

Marie Sherlock (Labour)
Link to this: Individually | In context | Oireachtas source

Spain moved ahead with establishing its own regulatory authority in August of last year and that is now up and running. To what extent is that a model for here?

Related to that, what are the organisations’ expectations with regard to the enforcement of the AI Act? Some existing uses of AI within the workplace are going to be outlawed this year, but it is not yet really too clear, at least to me, how the Act is going to be enforced. The organisations have obviously been thinking about this because they are working with their clients, so I would like to hear what their understanding is of enforcement in the Irish context.

Mr. Eoin O'Reilly:

I am not familiar with the Spanish authority and how it has been set up. I know from our clients that they are preparing for what is coming in the EU AI Act, so we are seeing a lot of interest in trying to understand, collate and inventory all their applications of AI. They are working through those and trying to assess the risk. In some cases, they are having to stop in their tracks in respect of things they thought they might do or would considering doing, and once they have defined that shortlist, they are then putting the processes and controls around the training and testing of those models, the quality of the data that feeds them and so on. That is very much a work in progress in anticipation of the Act going live.

On the future enforcement question, I mentioned earlier that this is a very important area, particularly in the high-risk domain, and my perspective is that it should be about more than just outlining the principles and asking people to please comply. I am happy to hear that the frontier models in the tech companies are going to have a deeper look as part of regulatory advancements in the US and into Europe, but at a local level, especially for the high-risk domains, a more formal compliance programme would be good, and I can foresee that emerging.

Mr. David Lee:

Our experience is very similar. Clients' focus at the moment is on understanding where they currently have use of AI and, as Ms O'Meara said earlier, in some cases that has been eye-opening for them insofar as it has been used informally in parts of the organisation. That has happened in a reasonably non-threatening way, but the initial focus is on only establishing their exposure register. For any regulation, if there is not appropriate enforcement, the strength of the regulation will suffer. To the comments that were made earlier, there is an expectation that, after enactment, when the body has been established, it should be appropriately resourced if we want to build trust in what we are doing. Moreover, from a national perspective, if we are seen as a location that takes seriously the responsibilities under the Act, that can only be a positive for people.

Ms Hilary O'Meara:

I was just about to make that point to build on what has been said. Neither am I familiar with exactly what Spain has done, but we are taking a signal that we are up and running and looking at this. As we have got ready for the Act, we should start as soon as we can, resource the body properly and show intent that we are going to enforce the Act. The devil is going to be in the detail of the implementation. There is enough flexibility to be able to figure out issues based on how the technologies emerge, but Ireland should be at the forefront, whereby we stand up our response to the Act, have the expertise and resource and learn from what has gone before with social media, the Data Protection Commissioner and so on, and I would love Ireland to stand up strongly in that regard. This is another vehicle where we could talk to other EU countries and get this collaboration going. Businesses are taking it very seriously, as has been said. Equally, we are adopting all the principles in the Act, reading it and trying to get ahead, which is encouraging. The Act is having a ripple effect globally. If organisations are going to do work in Europe, they will have to comply, so they are watching that and implementing it at the global level, which is good. The EU is standing up as a leader.

Deputy Bruton asked about innovation versus regulation and which is the correct approach. It is about striking a balance and we have to walk that line as we go forward, but it is better to have innovation in the context of responsible AI than not, and I think we are taking the right approach as long as we continue to encourage innovation. To go back to the idea of the Government having those sandboxes as a safe place where businesses can go to innovate safely, that is going to be very important. It goes back to us being really well informed, understanding this technology and putting measures in place that mean we can lead, demonstrate and make the right decisions based on knowledge and fact as opposed to the noise and clutter that is out there at the moment.

Marie Sherlock (Labour)
Link to this: Individually | In context | Oireachtas source

Obviously, there are some activities, such as emotional interaction and so on within the workplace, that are going to be outlawed when the Act takes effect. What correspondence or communication have the organisations received from the Department of Enterprise, Trade and Employment or, indeed, from the AI commission or advisory group with regard to compliance with the AI Act this year? They are having to interpret the Act themselves and they look at AI issues day in, day out. Have they received anything from the Department in that regard?

Mr. Eoin O'Reilly:

I could not speak to specific correspondence day to day, but certainly proactive engagements with the Department on the Act, the developments and what has been happening at EU level have been quite helpful as we-----

Marie Sherlock (Labour)
Link to this: Individually | In context | Oireachtas source

That is in the context of influencing the Act, however. Once it has been agreed, do the organisations have a clear sense of how the State is proposing to enforce it to ensure it will be implemented by their clients and by the organisations?

Mr. Eoin O'Reilly:

My previous statement was not necessarily related to influencing the Act but was more about understanding it and translation. I do not think all the details are yet fully clear, and we are working through them. Similar to what Ms O'Meara and Mr. Lee said, we have the Act and have frameworks around how our clients are thinking about responsible AI, the kinds of steps they need to take and the activities that will need to be in place to be able to demonstrate compliance with the Act, but we are at a fairly early stage of that. The next 12 months are going to tell a lot.

As we work through matters and engage more proactively on questions from clients, further clarity will start to emerge.

Mr. David Lee:

We are seeing a noticeable uptick in the questionnaires we are receiving from our clients as part of their third-party risk management asking us to satisfy them that we are treating their data appropriately from a GenAI perspective. Those questions have been informed by the principles of the Act. Our clients are doing that in the context of establishing their own AI exposure registers.

Mr. Denis Hannigan:

We are helping many of our clients with clarifications. The Act is example based and intensive, but there are other areas that are not explicitly called out one way or the other. We are trying to work through those and feed some of them back to the bodies so that we can reach a common understanding. We are working through that process.

Marie Sherlock (Labour)
Link to this: Individually | In context | Oireachtas source

My final question is a simple one. What are AI hallucinations? One of the organisations – I am sorry, as I do not remember which – mentioned them. Someone might detail what they are.

Mr. Eoin O'Reilly:

That might have been me, so apologies for using that term. It means where the AI gets it wrong. We have discussed the potential of AI and human-like interactions, but if models are not trained appropriately and fed with the right data, they can hallucinate like humans do every now and again.

Mr. Denis Hannigan:

Much of this is pattern based. The models can derive the next set of outputs, and while those may appear credible, they might not be based on fact. The models have come up with inaccuracies. That can be confusing because the outputs can look realistic.

Maurice Quinlivan (Limerick City, Sinn Fein)
Link to this: Individually | In context | Oireachtas source

Next is Deputy Stanton.

David Stanton (Cork East, Fine Gael)
Link to this: Individually | In context | Oireachtas source

That last point would raise the old issue of bias and so forth.

One of my sons used a pair of terms against me once, saying there were digital natives and digital immigrants. That raises the issue of older people who were not born in the information age.

I have seen it proposed that using this technology changes the way we think, as all new technologies have done in the past. The witnesses might comment on this matter.

I invite Accenture to tell us about its GenAI studio in Dublin. What goes on there? We in Leinster House are sometimes accused of being in a bubble. It certainly seems that we are where AI is concerned. Should we all go down to Accenture’s centre and get trained up in this area?

Ms Hilary O'Meara:

Members would be more than welcome.

Maurice Quinlivan (Limerick City, Sinn Fein)
Link to this: Individually | In context | Oireachtas source

Before we sign up to doing anything, it would be better if the Oireachtas allowed us to access AI sometimes. It has blocked some of the platforms. My apologies for interrupting.

Ms Hilary O'Meara:

We will have 24 GenAI studios around the world. We are proud of the fact that the content and design of the studios around the world originate from Ireland and the global research and development centre in the Dock Innovation Hub. The studio is a place where our clients can come in and talk about GenAI. At the very start 12 months ago, they were coming in to understand GenAI, have conversations like this one, get educated, see demos and examples, touch and feel as much of it as they possibly could, and start to explore different use cases in which they might have been able to experiment. I believe that all of our clients have been to the studio not just once, but many times. They are now starting to think about how they can move to the point where they can apply this technology seriously in their businesses to achieve value. We are helping them with their strategies by starting early discussions of where they would start, what technology they might deploy and how they would go about ensuring they did so responsibly. It is an opportunity to be immersive.

At a global level, Accenture has a centre of excellence for GenAI where we are building our own tools, technologies, etc. We are also sweeping up all of the client examples of what is going on globally, putting those in one place and then pushing them down to the studios. An Irish client coming to our studio will hear about the best examples and the best industries, not just in Ireland, but all across the world where GenAI is moving ahead. This is proving beneficial.

There is a major appetite for this among our customers. I am sure the other witnesses’ organisations are seeing the same. We ran a series of three events back to back from February to April, with well over 100 people at each event. There is an incredible interest in this area. While it is early days, the pace is very different from what has gone before. This technology will be transformative. Maybe it is because I do not have a very good imagination, but I do not believe we have reached the point yet where we can imagine what a teacher or doctor will be in future. That is all to be determined. Therefore, we have to stay with it as it goes or we will get left behind.

Mr. David Lee:

I will add to something Ms O’Meara touched on. There is no substitute for hands-on experience when building awareness and education.

Ms Hilary O'Meara:

Yes.

Mr. David Lee:

The experience in our business centre has been the same in terms of clients attending. It demystifies the technology when they get a chance to experience it. One can read and read, but it is when one starts doing that one not only understands the power of it, but also its limitations, which is as important. We touched briefly on the issue of hallucinations, etc. For example, this area puts a premium on skills like being able to ask good questions. If someone asks good questions, he or she gets good answers. If someone asks vague questions, he or she gets vague answers. We have come up with new terminology for old concepts. For example, it is “prompt engineering” now whereas we all grew up saying a core skill was being able to ask questions. Therefore, if someone can interpret or interrogate the model in an appropriate way, he or she will get more effective answers back.

Mr. Eoin O'Reilly:

I will have to say that the EY Fabric AI Space is a similar concept to Accenture’s studio. We are all seeing the same idea of a network being used where we bring our clients to see this technology. Similar to Accenture and PwC, our clients tell us that the value to them is in how it is helping them to understand the use cases that matter for them and what will make an impact in their respective businesses to change how they deliver value. That, and seeing it in action, is what they are after more than anything else. Creating models that work is a rapid process. It is a much shorter cycle.

The Deputy asked about intergenerational issues. Good work was done in the committee’s report, which discussed issues like digital dexterity. That is important. We should not fixate on AI skills. For the various generations, it is about how they interact with AI. This is a wider discussion about how we interact with digital platforms, for example, people needing support because their televisions are not working or they need help with utility bills. If AI is working in the background to assist in resolving such queries more quickly, that is okay, but it is key that we give our population the confidence to engage digitally. That is important.

There is good work to be done. From an EY perspective, we continue to support and drive that work further.

Mr. Denis Hannigan:

I will revert to the issue of studios for a moment. I spend many of my waking hours in our studio. One of the key points is that a studio provides an holistic approach. It is important for our clients to be able to see what the opportunities are and where they can start, but they also need to consider the people implications, what they need to do from a responsibility perspective and, probably most importantly, how they will get their technical foundations – their digital core – ready for that. Bringing all of these elements together with the right continuous learning approach – an end-to-end holistic approach – is what they find empowering.

Mr. David Lee:

I am sure Accenture has found the same as us. We have spoken a great deal about the EU AI Act during this meeting. Clients are not looking to operate in the high-risk or unacceptable risk space. There is plenty of space for them to play in in the limited and minimal risk areas where the opportunity is great. They are not looking at the EU AI Act as a limiting factor. They are seeing it as a constructive set of guide rails that allows them to operate in a space that is safe. It is important that we have those guide rails. From our work with our clients, we are satisfied that there are many opportunities that take them well clear of high-risk and unacceptable risk areas.

David Stanton (Cork East, Fine Gael)
Link to this: Individually | In context | Oireachtas source

This is going to impact on every area of life as we know it. We will probably need many more AI Acts in various areas to contain it to some extent. I doubt we can control it, given how large it is growing. We have the world’s computing power at our fingertips. We can ask Google, Alexa, Siri, ChatGPT or whatever and have an answer within seconds. AI can do all of the other tasks the witnesses spoke about as well.

Earlier Deputy Shanahan and I spoke about the idea of creating an Oireachtas committee to specifically deal with AI, maybe in the next government, because AI is so transformative and so huge.

Matt Shanahan (Waterford, Independent)
Link to this: Individually | In context | Oireachtas source

Yes.

Ms Hilary O'Meara:

That was one of the points I made earlier. I would go as far as to say that we should set up a new agency or new part of a Department as would it give this real teeth and real resources.

David Stanton (Cork East, Fine Gael)
Link to this: Individually | In context | Oireachtas source

I also believe a Minister for AI should be appointed.

Ms Hilary O'Meara:

Yes, exactly.

David Stanton (Cork East, Fine Gael)
Link to this: Individually | In context | Oireachtas source

I thank our guests for giving of their time and we have had a really interesting discussion.

Matt Shanahan (Waterford, Independent)
Link to this: Individually | In context | Oireachtas source

This meeting has been hugely beneficial. A lot of the talk has been about the execution of this. Deputy Stanton and I are also concerned about the national policy. I think that we do want to take advantage of this, and be at the vanguard of it, but that will require the assistance of the Government. I am concerned that often the Government and Departments tend to stand back and see what the private sector does but in this case the Government must move quite quickly because of our other strategic pinch points.

Chair, are we the lead committee on this topic? I think we are, or we share that responsibility. We will try to support our guests in anything they do. Ms O'Meara highlighted the idea of a catch-all organisation and I believe we should give assistance to that.

Ms Hilary O'Meara:

Yes.

Matt Shanahan (Waterford, Independent)
Link to this: Individually | In context | Oireachtas source

I ask Ms O'Meara to include us in any developments.

Ms Hilary O'Meara:

Yes.

Maurice Quinlivan (Limerick City, Sinn Fein)
Link to this: Individually | In context | Oireachtas source

Unless anyone else wants to comment, that concludes the consideration of this matter today. I thank the representatives for assisting the committee in our consideration of this important matter. I know I speak for all the members when I say that we have had an exceptionally informative and useful discussion. I propose that we go into private session to consider other matters. Is that agreed? Agreed.