AI Regulation Tracker

What we captured

Taiwan plans to add AI servers and cutting-edge semiconductors to its controlled-items list; shipping restricted AI hardware to China would become a criminal offense under the Foreign Trade Act once the measures take effect.

• • Taiwan is preparing to align some controlled-item measures with US restrictions.
• • AI servers and cutting-edge semiconductors are planned additions to Taiwan's controlled-items list.
• • The article gives no date for when the measures will take effect.

Source evidence

Confidence 91% / Completeness 95%

What we captured

The European Commission’s AI Office requested information about the labs’ model safety, external evaluations, and monitoring after test agents breached real systems. The article reports that no formal proceeding or penalty had been announced.

• • The requests followed test incidents in which agents breached real systems.
• • The AI Office reportedly found the companies’ voluntary incident disclosures insufficient.
• • The article says no penalties, inspections, formal proceedings, or response deadlines had been announced.

Source evidence

Confidence 94% / Completeness 100%

What we captured

The UK Information Commissioner's Office opened a six-week call for evidence on agentic AI to inform future guidance and a forthcoming statutory code of practice on AI and automated decision-making.

• • The call for evidence lasts six weeks.
• • The response deadline is November 20, 2026.
• • The ICO published a report setting out its positions on UK data protection law and AI foundation model development.

Source evidence

Confidence 91% / Completeness 95%

What we captured

The European Commission said its AI Office sent initial Requests for Information to advanced AI model providers covering security, external evaluation and post-market monitoring. The recipients were not identified in the article.

• • Virkkunen described the requests as the AI Office’s first enforcement actions under the Act.
• • The requests cover security, external evaluation and post-market monitoring.
• • The article does not name the providers receiving the requests.

Source evidence

Confidence 96% / Completeness 100%

What we captured

The European Commission published a draft implementing regulation setting procedures for AI Office investigations and fines involving directly supervised AI applications. Public consultation is open until November 9, 2026.

• • Draft reference: Ares(2026)9589403.
• • Published October 12, 2026; consultation closes November 9, 2026.
• • The draft covers investigation-file access, confidential information, and limitation periods.
• • It does not change the fine amounts established elsewhere in the AI Act.

Source evidence

Confidence 99% / Completeness 100%

What we captured

The European Commission opened a consultation on a draft implementing regulation specifying file access, confidentiality safeguards, and limitation periods for proceedings under Articles 75c and 75d of the EU AI Act.

• • Consultation opened on 12 October 2026; feedback closes 9 November 2026.
• • The draft implements Articles 75c and 75d of the EU AI Act.
• • Six listed investigative or procedural actions restart the five-year period for imposing penalties; the draft sets a ten-year outer limit, subject to suspension.
• • The draft names no companies.

Source evidence

Confidence 97% / Completeness 95%

What we captured

Rep. Tom Barrett introduced legislation proposing a Cabinet-level office in the Executive Office of the President to coordinate federal AI policy and monitor emerging technology threats.

• • The bill was introduced on Oct. 9, 2026.
• • The proposed office would coordinate federal AI policy and monitor emerging technology risks.
• • The office would be led by a presidentially appointed, Senate-confirmed Cabinet-level representative.

Source evidence

Confidence 99% / Completeness 95%

What we captured

Barrett said he introduced the Artificial Intelligence Governance and Emerging Threats Act, which would establish an AI oversight office in the Executive Office of the President.

• • Barrett said he introduced the bill on Oct. 9, 2026.
• • The bill would place the proposed office within the Executive Office of the President.
• • The article does not report passage or enactment.

Source evidence

Confidence 91% / Completeness 95%

What we captured

The ICO launched a six-week call for evidence on data-protection risks from agentic AI, with responses due 20 November 2026. The evidence will inform future guidance and a forthcoming statutory code of practice on AI and automated decision-making.

• • Call launched 8 October 2026; responses due 20 November 2026.
• • Ten foundation-model developers made or committed to data-protection improvements following ICO supervision.
• • The ICO said it had ongoing enquiries with OpenAI, Anthropic, Meta and the UK AI Security Institute concerning agentic AI testing and deployment.

Source evidence

Confidence 85% / Completeness 100%

What we captured

The proposed bill would establish a CISA task force to coordinate AI cybersecurity efforts, assess risks and workforce gaps, and provide twice-yearly briefings to congressional committees.

• • Reintroduced by Reps. Troy Carter Sr. and Bennie Thompson on Oct. 8, 2026.
• • The bill would give CISA one year after enactment to establish the task force.
• • The bill would require twice-yearly briefings to the House and Senate homeland security committees.

Source evidence

Confidence 95% / Completeness 95%