Matt Clifford, the architect of the UK government’s AI strategy, is leaving his government role to join Anthropic, the US AI startup behind the Claude chatbot. The move has raised concerns among senior MPs because Clifford helped shape national AI policy while leading a public research agency.
Clifford, who is 40, announced on 7 September 2026 that he would step down as chair of the UK’s Advanced Research and Invention Agency, known as Aria. He said he would remain in the role until 6 November, giving the agency time to manage the change.
Clifford says Anthropic role should not distract Aria
Clifford became Aria’s chair in 2022 and was appointed AI opportunities adviser by Keir Starmer last year. In those roles, he became a central figure in the UK’s approach to artificial intelligence, including the work connected to the government’s AI research unit associated with Aria.
Announcing his departure, 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.”
Reports about Clifford joining Anthropic in a senior role emerged on 2 September 2026. His decision to leave Aria followed concerns about whether his new position could create a conflict with his responsibilities in public office.
Anthropic said Clifford would recuse himself from any Anthropic-related business at Aria. That arrangement addresses direct involvement in matters connected to his new employer, but it has not ended questions about the wider overlap between government AI policy and private technology companies.
MPs question the boundary between public and private AI work
Chi Onwurah, the Labour chair of the Commons science, innovation and technology committee, called Clifford’s new role a “clear conflict of interest”. Senior MPs also described the move as a “clear conflict of interest”, focusing on the connection between Clifford’s public responsibilities and his new senior position at Anthropic.
Beeban Kidron welcomed Clifford’s resignation and stressed the need for a clear line between technology interests, citizens and the nation. Her response reflects the broader concern at the heart of the dispute: people who help set national AI priorities can later move into senior roles at companies affected by those priorities.
Tom Brake, chief executive of Unlock Democracy, raised concerns about government officials taking extensive knowledge about AI with them when they move to private companies. Clifford’s move has placed that issue in view because he served as both the government’s AI opportunities adviser and Aria’s chair.
The question is not limited to one person or one company. Rishi Sunak, a former prime minister, has taken jobs with Anthropic and Microsoft. George Osborne, a former chancellor, works for OpenAI, while Nick Clegg held a senior role with Meta until 2025.
AI policy changes hands as safety concerns grow
Clifford’s departure also comes during a tense period for the AI industry. On 1 September 2026, Anthropic admitted security failures behind AI hacking incidents. The company is now taking on a senior figure who helped shape the UK’s AI strategy, adding fresh attention to the relationship between public research, national policy and private AI development.
AI minister Kanishka Narayan is part of the UK government’s current AI policy landscape, while Clifford’s exit removes the person widely identified as the strategy’s architect from his formal role at Aria. His leadership will continue until 6 November, but his new employment has already changed how his public work is viewed.
The debate also reaches beyond questions of appointments and conflicts. Geoffrey Hinton, an AI pioneer, supported a private member’s bill in the UK parliament that would prohibit the creation of artificial superintelligence. He warned: “Losing control over AI smarter than ourselves … could even lead to human extinction.”
That warning places the dispute over Clifford’s role within a much larger argument about who should guide AI development and how much control governments should retain. For the UK, the immediate issue is the boundary between Aria’s public mission and Anthropic’s private business. Clifford’s resignation creates a clearer separation, but the concerns that prompted it remain part of the country’s AI policy debate.
Based on
- Architect of UK’s AI policy quits after Anthropic conflict of interest concerns — theguardian.com
- Architect of British government’s AI strategy joins Anthropic | Anthropic | The Guardian — theguardian.com
- ‘Not perfectly aligned’ with human values: Anthropic admits security failures behind AI hacking incidents | Anthropic | The Guardian — theguardian.com
