Matt Clifford, the chair of the UK government's "moonshot" research agency Aria, has announced he is stepping down after accepting a full-time job at Anthropic — a move senior MPs had called a "clear conflict of interest," according to The Guardian.

Clifford said on Monday that he would leave the chairmanship of the Advanced Research and Invention Agency, which funds high-risk frontier science and technology projects, to ensure his new role at the company behind the Claude chatbot "doesn't become a distraction." For more context on this story, see our ongoing AI trends.

From government adviser to Anthropic executive

The row began last week, when Clifford announced he was taking a full-time position at Anthropic leading the company's engagement with governments outside the United States — a portfolio that includes the UK government itself. At first, he planned to keep his Aria chairmanship alongside the job, with an arrangement agreed between Anthropic and the government under which he would recuse himself from any Anthropic-related business at the agency.

That compromise did not hold. Chi Onwurah, the Labour chair of the Commons science, innovation and technology committee, publicly labeled the setup a "clear conflict of interest" and wrote to AI minister Kanishka Narayan outlining her concerns.

"A clear conflict of interest"

Announcing his departure on Monday, Clifford said: "I have decided to step down to ensure my new role at Anthropic doesn't become a distraction from Aria's incredible work." He will remain in place until November 6, he said, "with appropriate safeguards against potential conflicts in place."

Onwurah welcomed the decision but said questions remained, including "how this situation arose, what conflict of interest assessments were undertaken and what safeguards are in place to protect confidence in Aria's governance."

Beeban Kidron, a crossbench peer and prominent critic of the government's AI policy, welcomed the resignation, telling The Guardian: "If we are to rebuild trust in the government, we need to see a clear line between tech interests and those of citizens and the nation."

Tom Brake, chief executive of the campaign group Unlock Democracy, had warned that moving extensive knowledge of AI policy into the private sector so soon after leaving government could weaken the UK's hand in pushing for a precautionary approach to AI development.

The revolving door between Whitehall and AI labs

Clifford's move drew attention because it fits a broader pattern. Former prime minister Rishi Sunak has taken roles with Anthropic and Microsoft since leaving office, while former chancellor George Osborne works for OpenAI. Both followed Nick Clegg, the former deputy prime minister, who held a senior position at Meta until 2025.

Clifford himself has been one of the most influential figures in British AI policy. The 40-year-old co-founded the venture capital firm Entrepreneurs First, became chair of Aria in 2022, and served as a key adviser to Sunak. He organized the 2023 AI Safety Summit at Bletchley Park and helped create the organization that became the world-leading AI Security Institute. In January 2025 he was appointed Sir Keir Starmer's AI adviser, stepping down six months later for personal reasons.

What Aria does, and why the chair matters

Aria was created to fund high-risk, high-reward frontier research of the kind that established agencies and universities often avoid, and its leadership is intended to be insulated from day-to-day politics. The chairmanship is a part-time public role, but it carries influence over which ambitious science and technology bets receive public money.

That is precisely why the overlap with a commercial AI role triggered alarm. Anthropic is one of the world's largest AI companies and an active participant in UK AI policy debates, from safety testing arrangements with the AI Security Institute to the siting of compute infrastructure. A chair who simultaneously leads Anthropic's engagement with the very governments funding and regulating the sector sat awkwardly with that separation, critics argued.

Monday brought no shortage of AI-in-government controversy: the same day, AI pioneer Geoffrey Hinton lent his name to a private member's bill being tabled in Parliament that would prohibit the creation of artificial superintelligence, warning that losing control of AI smarter than humans "could even lead to human extinction."

What it means for UK AI policy

The episode is likely to intensify scrutiny of how ministers police the movement of officials and advisers between government and the AI companies they regulate and fund. Recusal arrangements of the kind initially agreed for Clifford are a common fix, but critics argue they ask the public to trust that a part-time watchdog will remember to excuse himself — particularly when the employer in question is a frontier lab whose commercial fortunes depend on decisions taken across Whitehall.

For Anthropic, which is preparing for a reported mid-October IPO, the swift resolution removes a distracting controversy in one of its most important international markets. For the government, the pressure now shifts to answering Onwurah's outstanding questions about how the arrangement was approved in the first place — and to demonstrating that the safeguards promised until Clifford's November departure are more than window dressing.

---

Stay Ahead of AI

Get the latest AI news, analysis, and breakthroughs — all in one place.

Read more AI news →