AI Governance Watch

AI Governance Watch

Track AI governance worldwide: the news, the UN debates, every day.

A world of perspectives.

Follow the news.Read the Brief that goes to everyone.

Anatomy of a country record

A country, in context.

Latest news
The weekly Brief

Sent to everyone.

Every story, in order.

  1. UN & global forums·Six countries and major AI labs sign Paris Call declaration on frontier AI safety governanceCanada, France, Finland, Norway, the Netherlands and Bahrain joined Anthropic, OpenAI, Microsoft, Mistral and other labs in a joint declaration on AI safety oversight.Prime Minister of Canada
  2. Government·Finland and Estonia push EU for AI incident response systemAhead of the first European Council meeting on AI, Finland and Estonia call in a joint letter for EU-level arrangements to respond to serious AI incidents.POLITICO Europe
  3. Academia·Stanford's AI100 report finds AI governance outrun by capability, fragmented globallyStanford HAI's third AI100 study panel report warns international AI governance is fragmented and shaped by national security framing rather than cooperation.Stanford HAI
  4. Private sector·Censys finds more than 800 Artex AI hacking tool deployments exposed without a passwordThe open-source autonomous penetration-testing tool, linked to attacks on Korean banks and on ConnectWise, was reachable from the web without a password on more than 800 distinct deployments.techcrunch.com
  5. Civil society·Anthropic’s models are improving twice as fast, mostly due to AI, METR report findsThe Verge reports METR concluded after an 11-week investigation that Anthropic’s models are improving roughly twice as quickly as a year ago, that it is confident more training compute does not explain it, and that AI accounts for most of the acceleration; Anthropic was allowed to redact parts of the report.theverge.com

Said at the UN.

Every UN meeting, transcribed.

Every mention of AI is marked.

Find the AI passages. Read them in context.

Voted at the UN.

How countries voted on AI.

International Day of Safe, Secure and Trustworthy Artificial Intelligence

A/RES/80/302

144 in favour · 2 against · 3 abstaining · 44 not voting

See the full vote
UN & global forums

Six countries and major AI labs sign Paris Call declaration on frontier AI safety governance

Canada, France, Finland, Norway, the Netherlands and Bahrain joined Anthropic, OpenAI, Microsoft, Mistral and other labs in a joint declaration on AI safety oversight.

Prime Minister of Canada / 9 Nov

African Union and EU officials discuss AI governance at Addis Ababa security meetingAU and EU peace and security officials meeting in Addis Ababa on 6 November 2026 added AI governance to talks on conflict prevention and peacebuilding.
Time News · 6 Nov
AU and EU officials finalize AI governance cooperation plan in Addis AbabaSenior AU and EU officials met at the 9th Joint Retreat to finalize an implementation plan covering AI governance alongside maritime security and counter-terrorism.
allAfrica.com · 6 Nov
Lesotho urges child-centred AI governance at AU-EU peace and security retreatAt the 9th AU-EU Peace and Security Council joint retreat in Addis Ababa, Lesotho's AU envoy called for child protection safeguards in military AI use.
Africa Newsroom / Government of Lesotho · 6 Nov
Open Library

Today, everywhere

Thursday 12 November 2026 · 7:00 a.m. New York

Covering Wednesday 11 November 7:00 a.m. to Thursday 12 November 7:00 a.m.

Finland and Estonia urge EU arrangements to respond to serious AI incidents

Analysis by Ephemeris

Finland and Estonia called in a joint letter for EU-level arrangements to respond to serious AI incidents, ahead of the first European Council meeting on AI.1 The European Commission opened consultation on revised Digital Services Act election guidelines, urging platforms to test recommender systems and strengthen AI deepfake safeguards.2 Both moves seek EU-level answers to AI risks, while Stanford HAI's AI100 panel warns that international AI governance is fragmented.2,1,3

The panel's third report finds that AI governance has been outrun by AI capability.3 In India, the Madras High Court ordered the state anti-corruption agency to investigate a publisher accused of supplying suspected AI-generated books to a public library.4

Sources: Wednesday 11 November 7:00 a.m. to Thursday 12 November 7:00 a.m.

Key themes

One number

4

stories in this edition.

Coordinating AI governance across governments

Analysis

Finland and Estonia called in a joint letter for EU-level arrangements to respond to serious AI incidents, ahead of the first European Council meeting on AI. [1] Stanford HAI's third AI100 study panel report finds that AI governance has been outrun by AI capability. [3] The panel warns that international AI governance is fragmented and shaped by national security framing rather than cooperation. [3]

Coordinating AI governance across governments: stories

Elsewhere

Stories

No coverage today from: African Group, Asia-Pacific Group, Eastern European Group, Latin American and Caribbean Group, Western European and Others Group

Catching up

Older stories added in this edition’s window. Each shows when it happened and when it was added.

Catching up: stories

End of today’s edition

Next brief: Friday 13 November, 7:00 a.m. New York