Every AI Policy Ever · open in the interactive register

Published (non-binding, current)Non-binding · Guidelines / framework

Emerging processes for frontier AI safety

Record GB-02

Issued by
Department for Science, Innovation and Technology (DSIT)
Jurisdiction
United Kingdom
Level · region
National · Europe
Published / adopted
2023-10-27
Language
English

Summary

Overview of nine emerging safety processes and associated practices for frontier AI organisations, published by the UK government ahead of the AI Safety Summit of 1-2 November 2023 to inform companies' safety policies and help readers compare them. The processes are responsible capability scaling, model evaluations and red teaming, model reporting and information sharing, security controls including securing model weights, vulnerability reporting, identifiers of AI-generated material, prioritising research on AI risks, preventing and monitoring misuse, and data input controls and audits.

Safety topics

general-purpose / frontier AIevaluation & testingincident reportingtransparency & labellingcatastrophic risk

Status notes

Published with a companion page collecting frontier AI companies' safety policies, which the UK had encouraged companies to publish before the summit. The document describes itself as a snapshot that is not government policy that must be enacted and will need regular updating; six of its processes correspond to commitments many leading companies made in July 2023. GOV.UK shows it as published under the 2022 to 2024 Sunak Conservative government.

Sources

Open the official source https://www.gov.uk/government/publications/emerging-processes-for-frontier-ai-safety

Local copy

Archived. Last attempt 2026-10-17 20:42:01 UTC

Official source (third-party copy, not published) archive/GB-02/2026-10-11/official.html
Retrieved 2026-10-11 14:03:08 UTC
SHA-256 5677e79654a95eb7a7cc25e342e17ff547cca366cfaf183bd108075e231945a9
Linked document (PDF) (third-party copy, not published) archive/GB-02/2026-10-11/document-1.pdf
Retrieved 2026-10-11 14:03:11 UTC
SHA-256 32c4e01c6cb8bb5d2c16027dbcc4a4d4a5dac9a7f380a863c52da4a293593388

Verification

A
The instrument's own official text, or an official register, was stored and confirms the title, date and status. checked 2026-10-17 · official-url-fetched
Stored and read in this run: the GOV.UK publication page (last checked 17 Oct 2026, unchanged since 11 Oct) and the 45-page PDF. They confirm the title, DSIT as issuer, publication on 27 October 2023, the nine processes and the document's non-binding character; the page carries no withdrawal notice. Version 1's notes that it has not been formally replaced and was the forerunner of the Frontier AI Safety Commitments (INT-06) are not in these sources and were removed; the link to INT-06 stays in related.
Second check: Independent check by Claude Code (claude-sonnet-5-5), 2026-10-17: 5 field(s) supported.

History

  1. 2026-10-17notes: GOV.UK marks it 'historic' (published under the 2022-2024 Sunak government). It → Published with a companion page collecting frontier AI companies' safety policie — rewritten from the stored page and PDF; version 1's 'not formally replaced' and 'forerunner of the Frontier AI Safety Commitments' are not in this record's sources2026-10-17-recheck-europe-2 · Claude Code (claude-opus-5-5)
  2. 2026-10-17summary: A catalogue of emerging safety processes and practices for frontier AI developer → Overview of nine emerging safety processes and associated practices for frontier — made precise from the document: nine processes (version 1 omitted 'prioritising research on risks'), the summit dates and its purpose of informing and comparing company safety policies2026-10-17-recheck-europe-2 · Claude Code (claude-opus-5-5)

Related records

UK AI Safety Institute founding document United Kingdom · Superseded

Frontier AI Safety Commitments AI Seoul Summit (2024) · Published (non-binding, current)

Citation

Department for Science, Innovation and Technology (DSIT). Emerging processes for frontier AI safety. 2023-10-27. https://www.gov.uk/government/publications/emerging-processes-for-frontier-ai-safety. Status: Published (non-binding, current); verified 2026-10-17 (level A). Local copy archive/GB-02/2026-10-11/official.html retrieved 2026-10-11 14:03:08 UTC, SHA-256 5677e79654a95eb7a7cc25e342e17ff547cca366cfaf183bd108075e231945a9. Every AI Policy Ever, version 1.18 (snapshot 2026-10-18), record GB-02.

Improve this record

Know of an amendment, a status change or an error? Report news or an error on GitHub; the form opens with this record filled in.