Anthropic's long Whitehall reach

Plus ministers look at how to stem AI risks.

Share

Good morning,

Why are so many people who helped design Britain's AI policy now working for one of the companies at the centre of that policy? My WhatsApp lit up yesterday morning with less polite responses than that to the news Matt Clifford was taking a top job at Anthropic, while staying on as chair of ARIA.

For critics, it is evidence of an unhealthy amount of influence US AI firms have over the government. For defenders, it reflects the reality that government and industry must work ever closer to manage AI risks.

Both can be true, but having the country's moonshot research agency chaired by Anthropic's new international influencer-in-chief is a very odd look for ARIA. Why is the government happy to put an organisation it is pouring £1 billion into in an awkward position for the sake of one man's career?

It is also evidence of how Anthropic's relationship with the British state has moved well beyond that of a conventional tech supplier. I sketch out those links in today's main piece.

Happening Today 📆

In Birmingham: Reform's Party Conference kicks off at the NEC with "business day". More in Common hosts a panel on barriers to digital infrastructure investment at 12:30pm. Danny Kruger outlined the party's AI policies in July and he'll speak at 2:30pm tomorrow on "remaking the state". He'll also tell an IfG event at 1pm on Friday how Reform is preparing for government.

Parliament: MPs get their first chance at 9:30am to question the new ministerial team at the Department for Digital, Media, Culture and Sport. The questions range from grassroots sports to frontier AI. It's committee stage, day two, of the Cyber Security and Resilience Bill in the Lords.

While you're young: Public First has drinks for "YoPros" in tech. (I had to look up YoPros so I'm firmly ruled out).

Too many emails? Get updates on WhatsApp
Follow on WhatsApp