Ms Niamh Sweeney:

To avoid any confusion, we have 20,000 people working at Facebook on the wider safety and security of the platform, and just over 7,500 content reviewers.

The remainder of the 20,000 could be working on artificial intelligence, machine learning or various technical solutions to keep the platform safe. The Deputy is right that it is a big challenge because content review has never been done on this scale. On the need for direct State intervention, we receive millions of reports every week and if the State were asked to step in and do what we, Google and others do, it would be challenging to say the least.

There was a question on freedom of expression and whether it should be for Facebook to determine what should be left up and what should be taken down. That is an important part of the conversation. As Ms Quill said, we operate within the e-commerce directive in the area of illegal content and if we are made aware of illegal content on our platform, we have to act expeditiously to remove it. In other jurisdictions it is different, such as in Germany where the NetzDG Act covers illegal hate speech. In Ireland, it is left to us to discover what is illegal and many people have an issue with being left in a quasi-judicial position in this respect. We have a short period to determine what is illegal and there is a tendency to err on the side of caution, given the exposure companies have if they do not remove the content over time. If there is legal exposure for companies for pushing the boundaries on what constitutes freedom of expression, we could end up with the mass removal of perfectly legal content which may offend, shock or disturb but is not necessarily illegal.

On false accusations against public figures, we were part of the high-level working group on false news convened by the European Commission over the summer. This issue has travelled some distance since then and the focus now is more on electoral integrity. We have made a number of commitments as part of that and the Commissioner, Vra Jourová announced the agreed outputs in mid-October. We have undertaken to introduce additional tools with respect to electoral integrity this side of the European elections and we are working hard on this, which is in addition to the tool for viewing ads which we rolled out in April. This allows one to see any advertisement which an advertiser is running at the same time and addresses the idea of dark-outs.

If false accusations against public figures ever drift into defamatory territory, a clear process is in place. We have a dedicated reporting forum, as does Google, which is used to report defamation, which we know can be damaging. It is separate to reporting through our normal reporting channels, which are attached to every piece of content on the platform. We take this seriously.

Another question was on who should fund the office. I suspect that if it was funded by us it would attract criticism.