technology-policy6.385 min read
US Urges Meta to Open AI Models for Federal Safety Reviews
18 days ago•Source: The New York Times
Technology Policy News
The latest technology policy stories, summarized by AI
Featured Technology Policy Stories
View original source
3.635 min•26 days ago
Anthropic seeks White House resolution after export-control clash over Mythos 5
Anthropic is meeting with the Trump administration to resolve an export-control directive that forced the suspension of its AI models Fable 5 and Mythos 5 for all users by any foreign national, a move tied to national security concerns and a prior DoD supply-chain risk designation; the company describes the issue as a misunderstanding and is working to restore access while litigation continues.
More Top Stories
2
OpenAI pushes states-first path as White House tightens AI rules
Mashable•1 month ago
3
Executive Order Sets US AI Innovation and Security Path
The White House (.gov)•1 month ago
More Technology Policy Stories
Robocall Crackdown Could End Anonymity, FCC Signals Privacy Tradeoffs
The FCC is weighing sweeping Know Your Customer rules that would require government ID, a physical address, full legal name, and an existing phone number to obtain or renew voice service, with fines of $2,500 per illegal call charged to providers. While officials say the goal is to stop scams, critics warn the regime could erode semi-anonymous communications (burner phones) and sweep in red flags like crypto payments or nonmatching addresses, potentially harming vulnerable users; the rules are not yet in force and the commission is seeking public comment on privacy concerns.
White House AI policy chaos spooks Silicon Valley
The White House’s mixed signals on AI policy have left tech lobbyists anxious about potential rules to vet new AI models, with rival factions within the administration debating whether an executive order is needed and whether oversight should be voluntary (via CAISI) or mandatory; there’s no final plan yet, and industry voices urge clarity and a measured approach as frontier AI like Anthropic’s Mythos raises urgency.
US to pre-test Google, Microsoft and xAI AI models before release
The Commerce Department’s national standards agency will prerelease-test AI models from Google, Microsoft and xAI, with NIST evaluating their capabilities and security as part of expanded federal oversight of Silicon Valley; this follows a Biden-era agreement to pre-release testing by OpenAI and Anthropic, though concrete standards have not been set.
CAISI Expands Frontier AI Security Testing With Google DeepMind, Microsoft and xAI
The Center for AI Standards and Innovation (CAISI) announced new pre-deployment evaluations and post-deployment research agreements with Google DeepMind, Microsoft, and xAI to better assess frontier AI capabilities and security. The expanded collaborations, aligned with America’s AI Action Plan, position CAISI as the government liaison for AI testing and research, including testing in classified environments and ongoing interagency feedback through the TRAINS Taskforce. To date, CAISI has completed more than 40 evaluations to inform government understanding of AI capabilities and national security implications.
France tightens digital sovereignty with Linux-first government desktops
France’s DINUM ordered a Linux migration for its own desktops and mandated every ministry to draft autumn 2026 plans to cut dependence on non-European tech, expanding a sovereignty push that already includes replacing Teams/Zoom with a domestic platform (Visio) for 2.5 million civil servants by 2027. The move leverages La Suite Numérique and the Gendarmerie’s successful 103,000-seat Linux rollout as a governance model, while acknowledging open questions about software compatibility and the lingering dominance of non-European cloud infrastructure. The broader context is Europe’s bid for cloud and compute sovereignty amid strong US dominance.
California’s Age-Verification Push for OS Faces Feasibility and Privacy Hurdles
California’s AB 1043 would require OS providers to add an age-verification step and an API to relay the user’s age bracket to apps and stores by January 1, 2027, with penalties for non‑compliance. The piece flags questions about how practical and enforceable the measure is, potential privacy risks of collecting age data, and the impact on open-source projects and the broader app ecosystem, suggesting the law could be more nuisance than effective and may invite amendments or workarounds.
Tech Groups Rally Against Anthropic Supply-Chain Designation
Tech trade groups representing AI and software companies urged the Defense Department not to designate Anthropic as a supply-chain risk, warning such a move could disrupt AI workflows and chill U.S. innovation; letters argued disputes should be resolved through negotiations or procurement channels and cautioned the designation could set a broad, negative precedent for the U.S. tech sector, even as Trump administration actions to remove Claude continue and agencies proceed with the changes.
Capgemini pivots away from ICE amid French backlash
Capgemini said it will immediately sell its U.S. unit that works with ICE after French officials criticized the contract over immigrant surveillance; a Canadian company also canceled a sale to ICE, signaling growing political pressure on contractors doing business with the agency.