Biweekly AI policy review: 2026-07-01

[github-actions]

Biweekly AI policy review (2026-07-01)

Automated check from scripts/biweekly-ai-policy-check.mjs. Fires twice a month (1st + 15th). Triage and either:

• Add or update entries in the AI policy Worker registry, bump lastUpdated server-side, then close this issue, OR
• Comment noop and close if nothing actionable surfaced.

Current state of /api/ai-policy

• lastUpdated: 2026-04-30 (62 days ago)
• Policies tracked: 10
• By status: active (7), pending (2), repealed (1)
• By jurisdiction: United States (federal) (2), California (2), European Union (1), United Kingdom (1), China (1), South Korea (1), United States (voluntary, internationally referenced) (1), International (ISO standard) (1)

Policy-flavored news, last 14 days

Matched 19 articles (keyword scan; many will be tangential).

Date	Source	Title
2026-07-01	Hacker News AI	CorvinOS – self-hosted agentic OS where EU AI Act and GDPR compliance by design
2026-07-01	WIRED AI	Anthropic Added a New Security Measure to Get Back Into the Trump Administration’s Good Graces
2026-07-01	Hacker News AI	Multiple $20 AI Plans Are Better Than a Single $100 AI Plan
2026-07-01	ZDNet AI	Want a private ChatGPT alternative? How Proton's Lumo 2.0 locks down your data, EU style
2026-07-01	Hacker News AI	Platform Engineering 2.0 Mitigates AI Security and Compliance Risks
2026-07-01	The Verge AI	Google built a great smart speaker, but Gemini isn’t ready for it
2026-07-01	arXiv cs.AI	Investigating Multi-Agent Deliberation in Law
2026-07-01	arXiv cs.AI	Neuro-Bayesian-Symbolic Residual Attention Shallow Network: Explainable Deep Learning for Cybersecurity Risk Assessment
2026-07-01	arXiv cs.AI	AgentBound: Verifiable Behavioral Governance for Autonomous AI Agents
2026-07-01	The Verge AI	Anthropic’s long-sidelined Fable 5 is greenlit to return
2026-06-30	WIRED AI	The Trump Administration Is Lifting Its Export Controls on Anthropic’s Mythos and Fable AI Models
2026-06-30	The Verge AI	Libby will filter out AI content, kind of
2026-06-30	ZDNet AI	61% of US adults use AI for health information now - up from 2% in 2024
2026-06-30	MIT Technology Review	Building tech in the world’s secret R&D hub
2026-06-29	The Verge AI	Tidal won’t pay royalties on AI-generated music, but isn’t banning it outright
2026-06-29	NVIDIA AI Blog	Open Models, Closed Environments: Palantir Brings Secure AI to US Agencies With NVIDIA Nemotron
2026-06-28	The Verge AI	China’s Z.ai claims it can match Mythos on cybersecurity
2026-06-27	WIRED AI	Trump Administration Allows Anthropic to Release Mythos to Select US Organizations
2026-06-18	Hugging Face Blog	MosaicLeaks: Can your research agent keep a secret?

Sources: The Verge AI (5), Hacker News AI (3), WIRED AI (3), arXiv cs.AI (3), ZDNet AI (2), MIT Technology Review (1), NVIDIA AI Blog (1), Hugging Face Blog (1)

---

What justifies a registry update

1. A tracked policy hits a milestone (signed, enacted, partially applied, repealed).
2. A new framework lands at the federal or major-state level (EU member state implementation, US executive order, China measures).
3. A regulator publishes binding guidance that materially shifts compliance scope.
4. An enforcement action (fine, consent decree, lawsuit) establishes new precedent.

What does NOT justify a registry update

• Opinion columns about AI policy.
• Speculation about pending bills with low passage probability.
• Subnational guidance below state level (county / city ordinances rarely meet the bar).
• Industry self-regulation announcements (those are not policy in the registry sense).