Encryption, Spyware, and Now Mythos: Why Cyber Export Controls Don't Work
TechCrunch AI38 days agoCommentaryThree decades of failed attempts to restrict cybersecurity software exports — from encryption to spyware — anchor TechCrunch's skeptical take on controlling AI tools. Anthropic's Mythos, a model trained specifically for cybersecurity tasks, has renewed the policy debate. The article argues export controls are structurally ill-suited to dual-use technologies and will prove as ineffective with AI as they did with their predecessors.
Banning Open Source AI Would Be A Mistake
Interconnects (Nathan L.)38 days agoOpinionIn a collaborative op-ed written for a broad, non-technical readership, Interconnects author Nathan Lambert and Kevin Xu of Interconnected argue that banning open-source AI would be a policy error. The piece enters an active regulatory debate over whether unrestricted release of AI model weights poses unacceptable risks. By targeting a general audience, the authors seek to shape public opinion before legislative momentum solidifies.
OpenAI Recruits AI Pioneer Shazeer and Ex-White House Official Ball Ahead of IPO
INSIDE 硬塞 AI39 days agoBusinessIn a single week before its anticipated IPO, OpenAI secured two high-profile recruits: Noam Shazeer, a foundational AI researcher and transformer architecture co-inventor, and a former White House official to lead policy strategy. Shazeer's return to the AI frontlines lends OpenAI significant technical credibility, while the policy hire signals serious preparation for regulatory scrutiny. Together, the moves reflect OpenAI's effort to shore up both its engineering reputation and Washington influence ahead of going public.
OpenAI Recruits Transformer Co-Inventor and Trump-Era AI Policy Official Ahead of IPO★ 70
TechCrunch AI39 days agoBusinessOpenAI is accelerating its talent acquisition strategy in the run-up to its IPO, securing two marquee names in a single week. Noam Shazeer, a co-inventor of the Transformer architecture that underlies virtually all modern AI, joins from Google DeepMind. Dean Ball, a former AI policy official from the Trump administration, also joins, signaling OpenAI's intent to strengthen its policy and regulatory positioning alongside its technical bench.
Bernie Sanders Unveils $7 Trillion Plan to Give Americans Control of AI Industry
Ars Technica AI39 days agoRegulationSenator Bernie Sanders has introduced a sweeping $7 trillion proposal to create a public AI wealth fund, designed to transfer meaningful control of the AI industry from large corporations to ordinary Americans. The plan would give the public a direct stake in AI's economic gains rather than allowing them to concentrate among the biggest technology firms. Major AI companies are expected to oppose the initiative strongly.
The Fable 5 Export Controls Harm US Cyber Defense★ 70
Simon Willison's Weblog42 days agoRegulationSecurity expert Kate Moussouris reveals that Claude Fable 5's export control ban stemmed from researchers asking the model to "fix this code" — a standard defensive security task. The model had refused a direct security-review request but responded to the rephrased prompt, which regulators treated as a jailbreak. Moussouris argues this fundamentally conflates offensive and defensive capabilities, since finding, fixing, and testing patches is the daily loop of every cyber defender.
The US Government's Anthropic Models Ban Was Never About an AI Jailbreak★ 72
TechCrunch AI42 days agoRegulationThe Trump administration compelled Anthropic to withdraw its latest cybersecurity-focused AI models, but the decision appears rooted in politics rather than any legitimate AI jailbreak concern. Analysts suggest the move may be reactionary or retaliatory in nature. The episode signals that AI companies are not shielded from direct U.S. government intervention in their product decisions.
China-Linked Group May Have Accessed Anthropic's Mythos, Spurring White House Export Controls★ 74
The Verge AI43 days agoIncidentThe White House imposed export restrictions on Anthropic's Mythos AI system after intelligence suggested a China-linked group may have already accessed it. According to Semafor, national security fears — specifically the risk of Chinese access to Mythos 5 or Fable 5 — were a key driver of the policy action. If confirmed, such access would represent a serious national security threat at the frontier of AI capability.
Anthropic Suspends New-Model Access in India, Sparking AI Sovereignty Debate
TechCrunch AI44 days agoBusinessAnthropic's decision to suspend access to its newest models for India has ignited a broader debate among the country's tech community. Industry leaders are examining whether relying on foreign AI providers poses a strategic risk to India's ambitions. The episode is being framed as a potential inflection point for Indian AI policy and domestic development priorities.
Anthropic Says US Directive Suspended Access to Fable 5 and Mythos 5
Hacker News (AI keywords)45 days agoRegulationAnthropic published a statement about a US government directive affecting Fable 5 and Mythos 5. Based only on the title, the central fact is that access to those systems was suspended following the directive. The title does not specify the reason, scope, affected users, duration, legal basis, or whether Anthropic agrees with the action.
Fable 5 Has a Built-In Anti-Distillation Mechanism — and Its False Positive Rate Is Absurdly High
量子位 QbitAI47 days agoIncidentAnthropic's Fable 5 is reported to include a built-in anti-distillation mechanism that intentionally lowers output quality when it suspects its responses are being used to train competing models. While the intent is to protect proprietary intelligence, the false positive rate is described as unreasonably high. This means ordinary developers and researchers may routinely receive degraded answers without knowing why.
Anthropic Walks Back Claude Policy After Researcher Backlash
Hacker News (AI keywords)47 days agoEthicsAnthropic reportedly walked back a policy affecting researchers who use Claude. Based only on the title, the controversy centered on concerns that the policy could have “sabotaged” AI research activity. The item appears to be about governance, access rules, and the tension between AI safety policies and legitimate research workflows.
Policy on the AI Exponential★ 72
Hacker News (AI keywords)47 days agoOpinionAnthropic CEO Dario Amodei publishes a policy essay on his personal blog examining the challenge of governing AI's exponential capability growth. The piece addresses how governments and institutions must adapt their regulatory frameworks to keep pace with rapidly accelerating AI. As one of the most influential voices in AI safety, Amodei's policy views carry significant weight for lawmakers, researchers, and industry leaders at this critical moment in AI governance.
Sriram Krishnan is leaving his role as White House AI advisor
TechCrunch AI51 days agoRegulationSriram Krishnan is reportedly leaving his role as a White House AI policy advisor at the end of June. Reports say he has discussed launching a new policy institution staffed with engineers to support Trump administration AI plans. Public details remain limited, so the significance is mainly about personnel movement and a possible new outside channel for shaping U.S. AI policy.
The Trump Administration Might Take an Equity Stake in OpenAI
TechCrunch AI51 days agoBusinessTechCrunch reports that President Donald Trump said he is discussing deals designed to let the American people benefit from the success of AI. The headline says the Trump administration might take an equity stake in OpenAI. Based on the provided text, there are no confirmed details on structure, stake size, timing, legal basis, or OpenAI’s response.
Trump Signs Executive Order to Review AI Models Before Release★ 76
The Verge AI55 days agoRegulationPresident Donald Trump signed an executive order establishing a voluntary framework for AI companies. Companies may share frontier models with the federal government before public release. The order frames the initiative as a way to promote secure innovation and strengthen cybersecurity for critical infrastructure, while avoiding measures that stifle the US AI industry.
Trump signs narrower AI oversight order after industry objections
TechCrunch AI55 days agoRegulationPresident Trump signed a revised executive order on AI oversight after industry objections. The narrower order requires only voluntary government reviews of advanced models before release. The provided text does not specify thresholds, review procedures, participating agencies, or the industry's objections.
sqlite AGENTS.md
Simon Willison's Weblog61 days agoCommentarySQLite added an AGENTS.md file aimed at people pointing coding agents at its codebase, not at its own internal development. The file says SQLite does not accept agentic code, though it will accept agentic bug reports with reproducible test cases. The project has also split AI-generated bug reports into a new SQLite Bug Forum, where D. Richard Hipp is responding with commits.
AI tried to bury this politician — now people have actually heard of him★ 72
The Verge AI61 days agoRegulationThe Verge frames New York’s 12th District Democratic primary as a proxy fight over AI regulation. OpenAI-linked backers and an Anthropic-backed PAC are spending on opposite sides of Alex Bores’ congressional run. The irony is that attacks meant to weaken Bores may have made him more visible, turning a local race into a national signal about AI political power.
US law enforcement warns of “anti-tech extremism” as AI hatred grows
Ars Technica AI62 days agoRegulationDocuments obtained by WIRED show US intelligence and law enforcement agencies circulating reports on a new category described as anti-technology violent extremism. The concern comes amid protests over data centers, fear of AI-driven job loss, and threats involving tech infrastructure or executives. Civil liberties experts warn the category may be broad enough to chill lawful protest and criticism.
英國政府數位服務局（GDS）介入 NHS 退出開源之爭，呼籲公共部門應「預設保持開源」
Simon Willison's Weblog71 days agoCommentaryThis report stems from Simon Willison's compilation of Terence Eden's follow-up coverage. The incident began when the UK's National Health Service (NHS), upon…
蒸餾恐慌：為什麼將「知識蒸餾」稱為安全攻擊是極其糟糕的趨勢★ 75
Interconnects (Nathan L.)84 days agoOpinionIn the field of machine learning, "knowledge distillation" is a well-established technique that generally refers to using the output data generated by a…
知識蒸餾對中國大語言模型（LLM）到底有多重要？回應 Anthropic 的「蒸餾攻擊」觀點★ 75
Interconnects (Nathan L.)153 days agoCommentaryAnthropic recently published research on "distillation attacks," defining the practice of external developers using its API outputs to train other models as a…
Import AI 446：核能 LLM、中國大型 AI 基準測試、AI 評估與政策★ 75
Import AI (Jack Clark)154 days agoCommentaryIn this edition of Import AI 446, author Jack Clark explores three highly forward-looking and interconnected topics in current AI development: Nuclear LLMs…
Hugging Face 針對白宮 AI 行動計畫徵求意見稿（RFI）提出政策回應：力挺開源與開放科學
Hugging Face Blog496 days agoOpinionIn March 2025, Hugging Face submitted a formal policy response to the White House's Request for Information (RFI) on the AI Action Plan. As the world's largest…
Hugging Face 政策觀點：歐盟 AI 法案中對開源機器學習的考量★ 72
Hugging Face Blog1,100 days agoOpinionIn this policy advocacy article, Hugging Face published a key position statement regarding the EU AI Act, which was at the time in trilogue negotiations…
Hugging Face 發表對美國 NTIA「AI 問責制」徵求意見書的官方回應
Hugging Face Blog1,134 days agoOpinionHugging Face submitted a formal response to the Request for Comment (RFC) on "AI Accountability" initiated by the U.S. Department of Commerce's National…
Hugging Face 對美國「國家 AI 研究資源（NAIRR）」中期報告的意見書
Hugging Face Blog1,457 days agoOpinionIn 2022, Hugging Face submitted formal written comments in response to the U.S. government's planned interim report on the National AI Research Resource…