Yes. You have a number of recommendations in your report that will start to address the problem. We've talked a lot about the data piece: limiting their use of data, because that's where the value is coming from. As soon as you put a constraint on that, it restricts that profit.
There are very interesting developments, particularly in Europe, around copyright, and it seems they are moving forward again as of yesterday. I'd study those. They've been through a lot of the difficult balancing between copyright and free speech protection.
Yes, there are opportunities there. In your report, some of the evidence that Tristan Harris and a few others provided looked at ways to hold them liable for when they make recommendations. That is interesting to me, when we start to talk about not just links to the content but then using their AI to present recommendations. We have to actually put some liability on them. The real reason they're making money is that they have no liability for anything, but at the same time they get all the money. When there's a problem, it's not their problem; it's the Internet's problem or it's society's problem.