Peter Kyle

Peter Kyle became UK Secretary of State for Science, Innovation and Technology, heading the Department for Science, Innovation and Technology (DSIT), in July 2024. By January 2026 the post was held by Liz Kendall MP (Human Rights and the Regulation of AI (Joint Committee on Human Rights, HC 160, September 2026)). A Labour MP for Hove and Portslade, he was the principal public face of UK frontier-AI policy during his tenure, which since the Sunak-era Bletchley Declaration has narrowed in scope toward growth, AI diffusion, and targeted legislation rather than the broader horizontal "AI Safety Bill" expected earlier.

Background

Kyle became UK Secretary of State for Science, Innovation and Technology in July 2024 following Labour's general-election victory. DSIT was then the UK's lead department for frontier AI policy and housed the UK AI Safety Institute (AISI), described as the world's first dedicated government AI evaluation body.

Parliament's Joint Committee on Human Rights identifies Liz Kendall MP as "the then Secretary of State for Science, Innovation and Technology" when she set out an ambition in January 2026 for the UK to be "the fastest adopting AI country in the G7," and records that DSIT was abolished in July 2026, with AISI moving to the Cabinet Office (Human Rights and the Regulation of AI (Joint Committee on Human Rights, HC 160, September 2026)). The date on which Kyle left the post is not stated in the sources held here.

AI legislation

Kyle has repeatedly delayed the introduction of a comprehensive UK AI bill, initially expected in late 2024 or early 2025, while the government negotiates scope, preemption of existing regulators, and alignment with US and EU regimes. Government messaging shifted from an "AI Bill" framing to a narrower AI Safety Bill focused on frontier-model obligations, with other regulatory elements deferred to existing sectoral regulators. See UK AI Bill status source.

Kyle shepherded the Data (Use and Access) Act 2025 through Parliament. The Act is the UK's principal post-Brexit data-protection and data-use reform, relevant to AI training data, research access, and the UK's relationship with the EU GDPR regime.

AISI stewardship

During Kyle's tenure DSIT housed and funded the UK AI Safety Institute (which moved to the Cabinet Office when DSIT was abolished in July 2026), which remains one of the more technically respected national AI-evaluation bodies and a node in the international AISI network alongside the US AISI and IMDA Singapore. Kyle has publicly defended AISI's mission amid Treasury scrutiny and scepticism from some growth-focused Labour voices.

International engagement

Kyle represented the UK at the Paris AI Action Summit in February 2025. The United Kingdom, like the United States, declined to sign the summit declaration, citing concerns about its national-security language and governance provisions, which broke the alignment of the Bletchley and Seoul summits (Paris AI Action Summit Declaration (2025); UK AI Bill — Status and Delay (Source Summary)).

Positions and statements

Kyle has framed his approach as growth-first but not anti-safety, positioning the UK as an alternative to both US deregulation and EU prescriptive regulation, which he characterizes as "pro-innovation but proportionate," and citing AISI's technical capacity as the UK's distinctive asset. Critics across the spectrum describe the posture as insufficiently legible: safety advocates argue the AI Bill is being quietly shelved, while industry voices cite regulatory uncertainty.

Sources

• UK AI Bill Status
• Data (Use and Access) Act 2025 — Source Summary
• UK AI Safety Institute (AI Security Institute)
• Paris AI Action Summit Declaration (2025)

Relationships

• related: UK AI Safety Institute (AI Security Institute) — DSIT was the parent department until July 2026.
• related: Human Rights and the Regulation of AI (Joint Committee on Human Rights, HC 160, September 2026) — identifies Liz Kendall as Secretary of State in January 2026 and records DSIT's abolition.
• related: Data (Use and Access) Act 2025 — Source Summary — shepherded through Parliament.
• related: The Bletchley Declaration (AI Safety Summit, 1–2 November 2023), Frontier AI Safety Commitments (Seoul, 2024), Paris AI Action Summit Declaration (2025) — UK international AI diplomacy.
• related: JD Vance — both the UK and the US declined to sign the Paris declaration.
• instance-of: AI Safety Cases and Frameworks — UK AISI-centric model is a distinct national approach.