Background: Anthropic is the company behind Claude... and it's grown rapidly in the last few years. Founded in 2021 by ex-OpenAI staff who left over safety disagreements, it's now one of the biggest players in AI.

What happened: Over the weekend, Anthropic's CEO called for the AI industry to slow down model development, warning that AI could quickly run out of control. OpenAI and Grok's CEOs have also made similar warnings. So together, these big AI companies are proposing an AI slowdown. This would mean bringing in third-party model evaluators, shared safety standards across large language models (LLMs) and more coordination with governments. All the fun stuff.

What else: The interesting bit? Anthropic is planning to IPO at a valuation of nearly $2 trillion in the next few weeks. So some analysts reckon getting in and setting the standard for stricter AI development could give Anthropic and OpenAI an advantage. And that's where things start looking a little like regulatory capture.

What's the key learning?

💡 Sometimes the companies asking for tougher rules are also the companies best equipped to handle them. Compliance costs hit a tiny AI startup much harder than a company valued at $1 trillion.

💡 Regulatory capture is when industry leaders shape the rules that are meant to restrain them. Philip Morris, which owned 50% of cigarettes sold in the US in 2009, supported stricter cigarette regulation that banned sponsorships, colour ads and free giveaways, potentially making it harder for challengers to compete.

💡 Tougher rules can be good for an industry, but they can also entrench the leaders. Some experts think stricter AI standards could strengthen companies like OpenAI and Anthropic by making it harder for smaller players to keep up.