Risks to privacy are significant in the context of generative AI. That's certainly my view. It's the view of the G7 ministers.
I pointed to a recent report of the OECD in September of this year, in which the OECD canvassed all of the G7 digital and tech ministers about the top risks of AI—not just the risks, but the top benefits as well, because there are benefits in terms of productivity and so on. The top three risks in this report included privacy. The first one was disinformation and misinformation. There were risks to copyright. The risks to privacy were third, and you had risks of exaggerated biases and discrimination. Those are the top risks. I would agree with that categorization.