Cameron LUXTON

Well, yes, they've got in the uh fight a couple of months back, of course, with uh Donald Trump's uh uh uh administration in the wider Pentagon. They wanted to uh the US government wanted to use Anthropic and a bunch of other uh AI tools as part of their sort of defense and military operations. There was quite a disagreement over just what controls might need to be incorporated there. Uh depending on sort of which court you go to, it's either totally fine or totally not fine. Uh and so again, as is usual in the American system, they're going through the legal processes. But the the fact is anthropic's AI tools have become very much market leading. Uh, you've seen, particularly on their ability to uh support coding outcomes and and sort of co-work outcomes for non-technical uh operators. That sort of stuff really has uh just sort of blitzed the rest of the competition. You'll remember, of course, OpenAI when they came out with ChatGPT and similar, was driving a lot more activity in recent times. The open AI figures have started to drop back. They haven't been making uh quite as much money as had been anticipated. Anthropic's been moving forward a lot more than that, uh, particularly focused on enterprise software. But you're right, the big challenge at the moment is that all of this hype has come through, but now the sort of rubber is hitting the road. Where in particular, you've seen a lot of these AI companies that previously sort of just charged everyone a set fee per month. Now it's moving to more token-based pricing, and that sort of hit on the pricing uh front is starting to hit a lot of corporates. Uber, for example, uh spent its entire AI budget for a year in just four months, and basically had to say we can't spend anything more. So, with that sort of pricing tension starting to come in, there are questions around will these AI companies be able to make as much money uh money in the future, and that is raising questions around the eventual evaluat uh evaluation of these companies.