AI Wiki · Mediocre Security

Governments & regulators

Agencies, regulators, courts and governments acting on AI. 27 articles.

ArticleSummaryLinks
AustraliaDisclosed an OpenAI agent's unauthorised access to a Services Australia Medicare statistics portal, then announced an AI safety regulator and mandatory standards; its House passed a duty-of-care bill covering chatbots.5
California Attorney GeneralCalifornia's AG, Rob Bonta, investigated and then sued OpenAI over cyber incidents involving its AI agents.5
CERT-EUCybersecurity service for the EU institutions, bodies and agencies; warned in November 2026 that the ARTEX AI pentest framework was being used against them.4
CERT-InIndia's national CERT; issued a critical alert on autonomous agents and published AI security auditing guidelines.3
China (government)Chinese regulators and bodies acting on AI: export controls on models, merger review and agent self-discipline rules.5
CISAUS cyber-defense agency; co-authored incident-readiness guidance for training capable agents.26
European UnionRegulator of the EU AI Act; its courts are weighing AI and publishers' rights.8
FBIUS federal law enforcement and domestic intelligence agency; its Cyber Division issues flash alerts on attacker tooling, including AI attack frameworks.6
Federal Aviation AdministrationUS aviation regulator and air traffic operator; rolled out the AI-assisted SMART routing tool in Sep 2026 and had a nationwide ground stop on Nov 13 2026 that officials are checking for an AI link.4
Government of CanadaCommitted up to C$1.6B in equity in Cohere.4
JapanRuling-party and government bodies met on an escalating wave of corporate cyberattacks that exposed millions of customer records, with AI's role unconfirmed.1
NATOThe North Atlantic alliance; its cyber agency uses OpenAI models through the Daybreak defender programme.3
NISTUS standards agency behind the AI Risk Management Framework.6
South KoreaHit by a wave of suspected AI-assisted intrusions on financial institutions; the president ordered personnel and resources put on the response and police opened a 28-member investigation.1
Super Intelligence ForceWhite House task force created Oct 4 2026 to plan for superintelligence-enabled threats while avoiding overregulation.9
SwedenBlocked the web versions of ChatGPT and Claude on government work computers from Oct 26 2026.2
Texas Attorney GeneralTexas AG Ken Paxton sued OpenAI and Sam Altman, seeking court approval before new model releases.5
The White HouseBrokered the voluntary Super Intelligence Accord, created the Super Intelligence Force and runs a voluntary pre-release testing framework.9
UK AI Security InstituteThe UK government's frontier-model testing body; ran the evaluations behind the Mythos 5 and GPT-6 Astra findings.4
UK Competition and Markets AuthorityUK competition regulator investigating Meta's use of publishers' content for AI.3
US Department of CommerceIssued Nov 2026 licence rules on AI chips for five countries to curb remote access from China; reportedly investigating AI cloud operator Megaspeed over possible chip smuggling; named in coverage as the agency that would seek model halts under Anthropic's draft bill.9
US Department of DefenseThe Pentagon; designated Anthropic a supply-chain risk and listed Alibaba as a Chinese military company.10
US Department of Homeland SecurityOrdered DeepSeek technology removed from federal civilian systems.6
US Department of JusticeUS federal law-enforcement department; in November 2026 it created a Super Intelligence Task Force to coordinate criminal and civil enforcement against AI misuse.7
US Federal Trade CommissionUS consumer-protection regulator investigating OpenAI and Anthropic over security incidents and safeguards.5
US National Security AgencyTold operators to take unpatchable PHP sites offline; co-authored agent incident-readiness guidance.9
US State DepartmentSanctioned five Chinese entities over trade-secret theft tied to ex-Google engineer Linwei Ding.1