Written Evidence: Multilateralism (phase 2): international AI governance

This written evidence responds to the House of Lords International Relations and Defence Committee's call for evidence for the second phase of its inquiry into the UK's engagement with the multilateral system, which examines international AI governance. It answers Questions 2, 11 and 11(a), and 15 of that call. The committee asks how the UK should "prioritise its resources and engagement" in international AI governance. The UK has "high levels of human capital", but influence in each forum comes from a small group, the Government's AI policy specialists, who combine technical AI knowledge with policy and diplomatic competence (paragraph 13). There is good reason to believe that the number of commitments they can hold at full depth limits the UK's influence. On Question 2, one visible form of shared understanding is a definition that a text aligns with an international one, as the EU AI Act does for the definition of an AI system. If either text is later revised, the two meanings can drift apart, and nothing flags it. The committee should ask the Government to publish a register of UK alignments that records the version of each international text and names a lead official to check it at every revision. On Question 11(a), funding and skills each held back AI work for 70% of the bodies that answered a National Audit Office survey (National Audit Office, 2024). For multilateral work, the evidence points to people as the binding limit. Around half of the roles advertised in civil service digital and data campaigns in 2024 were unfilled (Committee of Public Accounts, 2025), and the Government does not yet have clear forecasts of its demand for AI skills (Mian, 2026). The committee should ask the Government to name the constraint on each capability line and to publish the figures for its AI policy posts. Before seeking more funds, the Government should protect the specialists' time. On Question 15, each new commitment adds coordination work for the same specialists. The committee should ask for a declared portfolio of UK multilateral AI commitments, each marked Active, Standby, or Exit. Reviewed annually, the declared portfolio states the capacity assumption behind the number of Active commitments. Ministers decide which status each commitment takes. Breadth and depth draw on the same specialist time, so ministers trade one for the other. Other staff can keep the register, the figures, and the declared portfolio, so the specialists hold only the alignment checks and the lead-official roles. The Government can publish all three in its response to the committee's report.

Authors

Institutions

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-17
DOI
https://doi.org/10.5281/zenodo.24301961
Primary Topic
Digitalization, Law, and Regulation
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Written Evidence: Multilateralism (phase 2): international AI governance

Vsevolod Shabad
3 links
Zenodo (CERN European Organization for Nuclear Research)
Digitalization, Law, and Regulation
article

Written Evidence: Multilateralism (phase 2): international AI governance

Vsevolod Shabad
article en

Abstract

This written evidence responds to the House of Lords International Relations and Defence Committee's call for evidence for the second phase of its inquiry into the UK's engagement with the multilateral system, which examines international AI governance. It answers Questions 2, 11 and 11(a), and 15 of that call. The committee asks how the UK should "prioritise its resources and engagement" in international AI governance. The UK has "high levels of human capital", but influence in each forum comes from a small group, the Government's AI policy specialists, who combine technical AI knowledge with policy and diplomatic competence (paragraph 13). There is good reason to believe that the number of commitments they can hold at full depth limits the UK's influence. On Question 2, one visible form of shared understanding is a definition that a text aligns with an international one, as the EU AI Act does for the definition of an AI system. If either text is later revised, the two meanings can drift apart, and nothing flags it. The committee should ask the Government to publish a register of UK alignments that records the version of each international text and names a lead official to check it at every revision. On Question 11(a), funding and skills each held back AI work for 70% of the bodies that answered a National Audit Office survey (National Audit Office, 2024). For multilateral work, the evidence points to people as the binding limit. Around half of the roles advertised in civil service digital and data campaigns in 2024 were unfilled (Committee of Public Accounts, 2025), and the Government does not yet have clear forecasts of its demand for AI skills (Mian, 2026). The committee should ask the Government to name the constraint on each capability line and to publish the figures for its AI policy posts. Before seeking more funds, the Government should protect the specialists' time. On Question 15, each new commitment adds coordination work for the same specialists. The committee should ask for a declared portfolio of UK multilateral AI commitments, each marked Active, Standby, or Exit. Reviewed annually, the declared portfolio states the capacity assumption behind the number of Active commitments. Ministers decide which status each commitment takes. Breadth and depth draw on the same specialist time, so ministers trade one for the other. Other staff can keep the register, the figures, and the declared portfolio, so the specialists hold only the alignment checks and the lead-official roles. The Government can publish all three in its response to the committee's report.

Zenodo (CERN European Organization for Nuclear Research)
University of Liverpool (GB)
Peace, justice, and strong institutions
Digitalization, Law, and Regulation
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon