Regulation & policy
Key Questions
What US government actions affected AI companies recently?
The US Commerce Dept forced Anthropic to suspend Fable 5 and Mythos 5 for three weeks due to a jailbreak vulnerability. OpenAI held White House stake talks valuing potential equity at $43B, and the Trump administration neared a voluntary standards deal with five labs.
What regulatory developments occurred at the state level?
Illinois Governor Pritzker signed SB315 into law, mandating annual third-party audits and whistleblower protections for AI. Maryland school districts face a fall deadline to set AI policies.
How are international regulators approaching AI governance?
Thailand's Draft Act introduces a risk-based approach with extraterritorial reach and data localization. The UK plans binding regulations for powerful foundation models with pre-market registration, and the EU AI Act compliance clock runs toward August 2026.
What proposals has Sam Altman made regarding global AI oversight?
In an FT op-ed, Altman proposed a US-led international forum modeled on the IAEA and suggested donating 5% of OpenAI equity to a US sovereign wealth fund.
What is the status of the EU AI Act in 2026?
High-risk provisions under the EU AI Act have been pushed to December 2027, with the overall compliance clock ticking toward August 2026 and no extensions expected.
How are courts and regulators viewing AI training data?
Courts and regulators are diverging on whether AI training constitutes fair use, creating uncertainty for model developers across jurisdictions.
What global warnings have been issued about AI safety standards?
A UN scientific panel warned that the global AI race has outpaced safety standards, while UK officials cautioned against waiting for an 'AI Hiroshima' before acting.
What public sentiment exists around AI regulation?
A bipartisan poll shows strong public support for mandatory AI safety reviews, and the Council of Europe advocates its AI Framework Convention as binding law.
US government intervention escalates: US Commerce Dept forced Anthropic to suspend Fable 5 and Mythos 5 for three weeks due to a jailbreak vulnerability, sparking rally in decentralized AI tokens. OpenAI's White House stake talks put a $43B question over AI power. Trump administration nearing voluntary standards deal with five labs (Meta holding out), including 30-day government review window and jailbreak severity scoring. FTC policy statement targets 'ideological steering' as consumer deception, challenging Colorado's anti-discrimination law. Sam Altman's FT op-ed proposes US-led international forum (IAEA analogy) and 5% OpenAI equity donation to a US sovereign wealth fund. New academic paper proposes Sovereign Shareholder Firewall. Brookings argues G7 should accept companies' standards offer but make it enforceable. FIRE's deep dive into AI free speech battles. RAND report on European table-top exercises. Illinois SB315 signed into law, mandating annual third-party audits and whistleblower protections, setting a new state-level bar. Bipartisan poll shows strong public support for mandatory AI safety reviews. UN scientific panel report warns global AI race has outpaced safety standards; July 6-7 Geneva dialogue. Thailand's Draft Act introduces a hybrid model with risk-based approach, extraterritorial reach, strict liability, and data localization, adding a non-Western perspective to global regulation. Courts and regulators diverge on AI training as fair use. Comprehensive global regulation roundup for 2026 highlights EU AI Act high-risk pushed to Dec 2027, US federal pushback against state laws. UK's Yvette Cooper warns 'AI Hiroshima' cannot wait. Dongfang Suanxin exits stealth with 3D stacking chips to bypass US export controls. Perplexity co-founder and Yann LeCun challenge safety-first narrative as power grab. Anthropic urges Congress to require mandatory AI safety tests with compute/revenue thresholds and preemption of weaker state laws. EU AI Act compliance clock ticks toward August 2026 deadline with no extension. Legal guide clarifies CSAM reporting obligations for AI platforms. Free-market critique of OpenAI government equity proposal adds to debate. State-by-state AI regulation reference (7 in effect, 3 enacted, 20 proposed). Guterres warns 'last generation' to shape AI rules. FCA official calls for regulating LLMs as general-purpose AI tools. Council of Europe AI Framework Convention advocated as binding law. Maryland school districts face fall deadline for AI policies. UN Global AI Governance Dialogue launched in Geneva, Guterres warns 'last generation' to shape rules, scientific panel warns safeguards lag capabilities. New: UK plans binding regulations for powerful foundation AI models, with pre-market registration proposal, drawing on Alan Turing Institute analysis. New: Illinois SB315 detailed breakdown confirms annual third-party audits and whistleblower protections, setting a new state-level bar. New: Thailand's draft AI Act adopts a risk-based approach with extraterritorial reach, strict liability, and data localization, adding a non-Western regulatory model. New: Virtual Assets Regulatory Authority CEO argues AI adoption in finance bottlenecked by regulatory infrastructure, not technology.