My Lords, I agree with the noble Baroness, Lady Kidron, that all these amendments are very much heading in the same direction, and from these Benches I am extremely sympathetic to all of them. It may well be that this is very strongly linked to the categorisation debate, as the noble Baroness, Lady Kidron, said.
The amendment from the noble Lord, Lord Bethell, matters even more when we are talking about pornography in the sense that child safety duties are based on risks. I cannot for the life of me see why we should try to contradict that by adding in capacity and size and so on.
My noble friend made a characteristically thoughtful speech about the need for Ofcom to regulate in the right way and make decisions about risk and the capacity challenges of new entrants and so on. I was very taken by what the noble Baroness, Lady Harding, had to say. This is akin to health and safety and, quite frankly, it is a cultural issue for developers. What after all is safety by design if it is not advance risk assessment of the kinds of algorithm that you are developing for your platform? It is a really important factor.
6.45 pm
I hope that we adopt that in AI regulation more broadly than simply online safety of this kind for social media platforms. We need a change of culture so that this is not just a question of developing without thinking about the ethical aspects of it. It is really important that we start with this kind of debate talking about assessing risk upfront. That should be the key test and not the size or capacity of a particular platform.
I support these amendments. I hope the Minister can give us some indication that we are all heading in the same direction as he is or that he is heading in the same direction as us. That would be enormously helpful.