Anthropic上市前聘请英国游说公司应对AI反弹Anthropic Taps UK Lobby Firm Before IPO to Counter AI Backlash

PBC hired a lobbying firm run by the one-time aide to former British prime minister Boris Johnson, adding to the ranks of UK political veterans the company has recruited to handle mounting calls to regulate artificial intelligence.

The Claude developer retained Charlesbye Ltd., a firm managed by Lee Cain, Johnson’s former communications director, during the most recent quarter, according to filings in the UK lobbyist registry. It’s the first time Anthropic has appeared in the filing, where lobbyists list clients who pay to contact a minister or permanent secretary.

Anthropic is readying a massive initial public offering at an uneasy political moment for tech. Silicon Valley’s largest AI developers are expanding their global political operations while there’s rising public concern with the risks of the technology, its impact on jobs and its related data center footprint. Last year, OpenAI hired former UK Chancellor George Osborne to lead international efforts.

But Anthropic has gone further, recruiting a wave of older Downing Street figures, including former premier Rishi Sunak and Liam Booth-Smith, his chief of staff. Matt Clifford, an AI adviser to Sunak’s successor, joined Anthropic in September. The company hired at least nine more veterans of the UK’s Treasury, tech and Cabinet departments, according to LinkedIn profiles.

thropic spokesperson said that the company supports the UK government’s efforts “to understand and manage the risks and benefits from power AI.”The company didn’t comment further on the Charlesbye contract or how many people it has hired from government.

Cain declined to comment. His firm markets itself as providing advice on corporate strategy, crisis management and public affairs.

Much of Anthropic’s future growth relies on having enough computing resources at data centers to train and run its AI models. Yet after the US midterm elections, a wave of newly elected governors are planning to stop or pause data center construction. The public backlash has extended to Europe, where officials have pressed for alternatives to American software for sovereignty reasons.

The UK has also yet to decide how to legislate AI. Britain’s ruling Labour Party promised “binding regulation” of AI model developers in its election manifesto, but it has not passed such a measure. AI minister Kanishka Narayan struck a softer tone after he was named to the cabinet in July, pledging that the government would “legislate wherever necessary” and commit resources to dealing with dangers like cybersecurity attacks.

UK officials are looking to add new rules around AI agents and chatbots. Prime Minister Andy Burnham has signaled he wants closer ties with the EU, which has implemented broad regulation on the tech.

Anthropic is one of the most vocal companies about the risks of powerful AI models. “We have consistently believed that there should be binding rules on companies and that voluntary commitments are not enough,” Booth-Smith, who was hired to run Anthropic’s policy efforts across Europe, told a UK parliamentary committee last month.

After leaving Downing Street in 2024, Sunak joined as a senior adviser to Anthropic, a role he likened to serving as “an internal think tank.”While in office, he hosted an AI safety summit and set up the AI Security Institute, a government agency that evaluates leading models, including those from Anthropic. Booth-Smith has said that Anthropic researchers communicate weekly with staff from the UK institute.

In the first half of the year, Anthropic had 11 meetings with UK government officials, according to fillings from four departments. OpenAI met five times with officials during the same period.

The Anthropic spokesperson said that Sunak and Booth-Smith do not participate in “any lobbying of the UK government.”Charlesbye is a relatively small firm, but it has strong links to the Tory party. Helen MacNamara, a former deputy cabinet secretary, recently joined its advisory board. Another director served as an adviser to Sunak. On its website, the firm lists Amazon.com Inc., BT Group Plc and the Premier League among its clients.

Earlier this year, Anthropic hired the US lobbying firm Ballard Partners after the Pentagon designated the AI firm as a supply chain risk following a dispute over using its tech in weapons systems.

In addition to the UK, Anthropic is currently searching for community managers for its data center efforts in the US and Australia, as well policy positions across the US and in South Korea, according to its hiring page.