AI Policy Wiki
Dashboard
⚲
AI Policy Wiki › People & Organizations

Kanishka Narayan

high confidence · updated 2026-09-23

UK Labour MP for the Vale of Glamorgan and Minister of State for Artificial Intelligence since July 20, 2026 — a joint Cabinet Office and Department for Business, Innovation, Science and Trade post created when DSIT was abolished. Previously Parliamentary Under-Secretary of State for AI and Online Safety from September 2025.

Contents

Kanishka Narayan is a British politician who has served as the United Kingdom's Minister of State for Artificial Intelligence since July 20, 2026. A Labour member, he has been the Member of Parliament for the Vale of Glamorgan since July 2024 (Source: gov.uk). He is the first UK minister to hold artificial intelligence as a dedicated cabinet-level brief.

Background

Narayan was born in November 1989 in Muzaffarpur, Bihar, India. Before entering Parliament he was a civil servant in the Cabinet Office under David Cameron and later worked at the environment department under Liz Truss (Source: en.wikipedia.org). He moved from Cardiff to Barry, in the Vale of Glamorgan, in 2023, weeks after being selected as the Labour candidate for the seat, and won it at the July 2024 general election, unseating Alun Cairns with a majority of 4,216 (9.2%). He is Wales's first ethnic-minority MP (Source: en.wikipedia.org).

Ministerial roles

Narayan was appointed Parliamentary Under-Secretary of State for AI and Online Safety on September 7, 2025, a junior ministerial role within the Department for Science, Innovation and Technology, succeeding Feryal Clark. He held that post until July 20, 2026 (Source: gov.uk).

On July 20, 2026 he was appointed Minister of State (Minister for Artificial Intelligence) in the reorganization under Prime Minister Andy Burnham that abolished DSIT, and was elevated to cabinet (Source: gov.uk; techieray.substack.com; bbc.com).

The joint-department structure

The post is unusual in being held jointly across two departments, which splits the AI brief along an industrial-versus-governmental line. Narayan's portfolio in the new Department for Business, Innovation, Science and Trade covers AI policy, data infrastructure, AI Growth Zones, the UK AI research resource, AI for science, robotics, semiconductors, Sovereign AI, and AI innovation and economy-wide adoption. His Cabinet Office portfolio covers AI strategy, the AI Taskforce, the AI Security Institute, the AI Economics Institute (jointly with HM Treasury), and public-sector adoption (Source: gov.uk).

That division answers the question left open by initial reporting on the reshuffle: the former DSIT AI portfolio was not transferred wholesale to a successor department. Compute, semiconductors, sovereignty and industrial adoption went to BIST, while the evaluation and strategy functions — including sponsorship of the AI Security Institute — moved to the centre of government in the Cabinet Office. See Sovereign AI and AI Sovereignty.

Activity in the role

Announcements issued under Narayan's name in his first six weeks indicate where the portfolio's weight has fallen. They include a £100 million competition backing British AI companies to work on public services (August 31, 2026); practical AI training for residents of Barnsley to support employment (August 25, 2026); a £300 million investment in the Lanarkshire AI Growth Zone, with Dell establishing a Scottish base (August 18, 2026); a Sovereign AI investment in a UK startup working on AI chips (August 3, 2026); and the appointment of Lord Vallance as chair of a new Prime Minister's AI Taskforce (July 24, 2026) (Source: gov.uk).

In his previous junior role he travelled to Australia in May 2026 to study that country's social-media approach, and the two governments concluded a pact on fast-moving AI security risks announced May 25, 2026 (Source: gov.uk).

Narayan attended King Charles III's meeting with AI executives at Dumfries House on September 17, 2026 (Source: reuters.com). See King Charles III.

Narayan accompanied Prime Minister Andy Burnham to New York in September 2026, where Burnham said on September 21 that the United Kingdom would use its G20 presidency to seek a global agreement on AI, acting as an "honest broker" between the EU, the United States and China (G20 (Group of Twenty); Source: politico.eu).

Positions and statements

Narayan gave oral evidence to Parliament's Joint Committee on Human Rights on 25 February 2026, as Parliamentary Under-Secretary of State and Minister for AI at DSIT, and wrote to the committee on 1 April 2026. The committee's report of 14 September 2026, which describes his previous post as including the title Minister for AI Opportunities and notes that he "will attend Cabinet" in the new role, records the following positions (Human Rights and the Regulation of AI (Joint Committee on Human Rights, HC 160, September 2026)):

  • Human rights and the Framework Convention. "[A]s a country, we want to make sure that AI is grounded in human rights, democracy and law." On ratifying the Council of Europe Framework Convention, he said "I feel no significant obstacles in us being able to put [the Convention] into effect," but that the programme of work required "is pretty extensive."
  • Regulatory approach. "[S]tatic forms of regulation, where we have been historically, may not be best adapted to where we are likely to go." He said the Government carried out systematic reviews to identify risks not being effectively countered and acted to close gaps.
  • The AI Security Institute. The institute "has some of the world's most exceptional talent in being able to assess models" and "is unparalleled globally in its access to models pre-deployment"; he had seen no indication of a need for reform to its access and capabilities but would keep the question under review. He said its 2025 renaming may have helped it "attract particular types of talent," and that it coordinates with Ofcom to ensure Ofcom's functions are "appropriately deployed upstream." His correspondence is the committee's source for the statement that AISI has no regulatory authority and cannot demand model access or prevent release.
  • Public-sector safeguards. Public-sector AI use involved "very specific focus on ensuring compliance with the equality duty and avoiding any discriminatory outcomes," and new guidelines for AI-ready government datasets "require human oversight at key stages to protect privacy rights, a duty to prevent bias," and human accountability for decisions.
  • Automated decisions. Automated processing should be used only where "proportionate and fair," and in a way that "still allows individuals to be informed, to contest, and to secure human intervention."
  • Online harms. The Government was taking self-harm and suicide content online and in AI chatbots "extremely seriously."

The committee welcomed the increased prominence of AI in government under his expanded role but reached different conclusions from the approach he described, recommending a cross-sector AI Bill, a single statutory AI oversight body and a statutory AISI with mandatory pre-release submission of powerful models (Human Rights and the Regulation of AI (Joint Committee on Human Rights, HC 160, September 2026)).

Open questions

  • Ownership of the delayed UK AI Bill after the DSIT reorganization is not stated on the sources reviewed here; the gov.uk portfolio listing names neither the Bill nor primary legislation.
  • The relationship between the new Cabinet Office AI Taskforce chaired by Lord Vallance and the AI Security Institute's existing evaluation remit is not established here.

Relationships

Categories: People & Organizations | uk | government | agency | ai-policy | compute | safety
What links here (13)