In April last year, two young men died by suicide. I am not going to put their names on the record but they are in the public domain. Both of these young men had been chatting with an AI bot at the time of their deaths. Their families say that a specific version of this artificial intelligence bot designed to be friendlier, more empathetic and encouraging had in fact encourage the deaths by suicide of these young men. These claims made by the families are being investigated. These cases serve as a lesson because AI is easy and fast, and it is becoming a tool more and more people rely on. Therefore, it will increasingly become a tool that people communicate with. It is not necessarily always a force for good and there are more than two families who have real reasons to watch this and other debates as they try to make sure we get this right. We know that artificial intelligence offers huge potential but it carries with it great risk, as is the case with all new technologies. There has to be regulation. This particular form of technological endeavour might move a little faster than others, but that does not mean regulation should not be able to move fast as well. We know the EU has put a lot of store in its digital policy work. It obviously wants to be a world leader in the online space, as does Ireland because that would make us good and fair Europeans. In practice, it is going to fall to Ireland, as the place where many of these companies are based, to be the regulator. In the previous Dáil, I was worn out pleading with the then Minister for enterprise, Leo Varadkar, to intervene and help content moderators, as their jobs were being outsourced and downgraded. Content moderators are the front-line defenders when it comes to online activity. They are the people who stand between the user and the AI platform. The then Minister response was that a lot of that work would probably be done by AI in the future. Mr. Varadkar is not here but I am sure if he was, he would regret those words. We need to put safeguards in place, which will not be easy. This cannot just mean banning social media for under-16s because social media can be harmful to over-16s and over-18s. Austin Gordon, who was 22 when he died by suicide, had identification that would have enabled him to use the platforms. The protections must be real and workable.

Sentiment score: 0.09