News-June-29-2026

June 29, 2026 — AI law and policy developments included U.S. cybersecurity gating of OpenAI and Anthropic frontier models, an Austrian request that the European Union explore hosting Anthropic inside the bloc after U.S. access restrictions, and California's statewide-local agreement for discounted Claude access.

Contents

1. OpenAI and Anthropic limit frontier model access during U.S. cybersecurity review
2. Austria asks EU to explore hosting Anthropic after U.S. model-access curbs
3. California makes Claude available to state and local government agencies

---

OpenAI and Anthropic limit frontier model access during U.S. cybersecurity review

OpenAI restricted the rollout of its GPT-5.6 model series at the request of President Donald Trump's administration while the government conducts a cybersecurity review. Help Net Security reported that OpenAI began rolling out GPT-5.6 in limited preview to trusted partners and said the release was being coordinated with the U.S. government before broader availability through ChatGPT, Codex, and the API. AP News reported that Anthropic also limited access to a new model during the review process. The development is notable because it frames frontier-model release timing as a cybersecurity and national-security review issue rather than only a commercial deployment decision.

---

Austria asks EU to explore hosting Anthropic after U.S. model-access curbs

Austria's state secretary for digitalisation, Alexander Pröll, asked European Commission technology-sovereignty official Henna Virkkunen to explore the "strategic establishment and participation of Anthropic within the European Union," according to The Next Web's account of the letter. The request followed U.S. Commerce Department restrictions that required Anthropic to cut off foreign nationals' access to its two most advanced systems, which The Next Web reported led Anthropic to switch the models off for users outside the United States because shared cloud access could not be separated by nationality. Pröll framed a possible European presence for Anthropic as a way to offer legal certainty, market access, capital, and European values while reducing dependence on U.S. policy decisions for access to frontier AI systems. The proposal raises legal, competition, security, and AI-sovereignty questions for the European Union because it would rely on hosting an American AI company rather than only building a domestic European champion.

---

California makes Claude available to state and local government agencies

California Governor Gavin Newsom announced that California entered a partnership with Anthropic to make Claude available to all state agencies as well as cities and counties. The Governor's office said the agreement gives state agencies access to Claude at a 50% discounted price, with free workforce training, generative-AI technical assistance, and workflow input from Anthropic developers. The same discounted offer also applies to California local governments, including cities and counties. TechCrunch reported that the agreement gives state agencies and local governments access to Claude, training, and support from Anthropic at a time when organizations face high enterprise-AI subscription costs.