Google DeepMind 正与英国政府、Google Cloud、Faculty 及 Barnet、Dorset、Camden 地方规划部门合作，基于 Gemini 共同开发 AI 规划原型工具，目标将住户规划申请审批时间缩短 50%。

The US Commerce Department issued an export control order last Friday requiring Anthropic to block all foreign nationals—including its own non-US employees—from accessing Mythos 5 and Fable 5. Anthropic fully disabled both models and sent executives to Washington to negotiate with Treasury Secretary Bessent and Commerce Secretary Lutnick. Anthropic argues the jailbreak cited by the government is narrow and non-universal, and that OpenAI's GPT-5.5 can achieve the same capability. Amazon CEO Andy Jassy may have reported red-team findings to the government, but Anthropic says the same conclusion holds for GPT-5.5. The post doesn't disclose the status of negotiations or when the models might return.

Anthropic abruptly took its newest Fable 5 and Mythos 5 models offline over the weekend at the White House's request. The US government demanded it block access for all foreign nationals, including its own employees. The incident is a blunt reminder that the US not only dominates frontier AI—its government can also decide who gets to use it. The post doesn't spell out how long the shutdown will last or what specific safeguards were already in place.

Dozens of cybersecurity experts urged the White House to lift export controls on Anthropic's Fable and Mythos models. They argue the ban will limit defenders' ability to secure software and products. The post is an RSS snippet—it doesn't name the signatories or include a White House response.

Anthropic has cut off access to its Mythos model following a US government order. The body is a Bloomberg video report; it does not disclose the order's legal basis, specific rationale, or shutdown timeline. What's confirmed: Anthropic complied, and Mythos is no longer available. Wait for the written order or an Anthropic statement before drawing conclusions on scope and precedent.

India signed a deal with UAE-based G42 to deploy a Cerebras-powered AI supercomputer on Indian soil. It gives India a second path beyond renting from AWS, Microsoft, and Google: own the hardware, keep data under local governance, and work with a non-US operator. India is the first country to join G42's 'Intelligence Grid.' Financial terms and hardware ownership are not disclosed. The real test is whether G42 can match the software and developer ecosystem of the US cloud trio.

Anthropic's top models Mythos and Fable were hit with export controls over safety concerns and taken offline. The company is now flying senior technical staff to D.C. to meet White House officials and patch things up. Both sides say they want a quick resolution, though the administration had complained Anthropic wasn't engaging seriously.

OpenAI confirmed Saturday that a coalition of states including New York and Colorado subpoenaed the company Friday, seeking internal documents on user data handling, minor safety, and advertising. OpenAI said it takes the concerns seriously and noted the latest ChatGPT version adds safeguards like parental controls. The probe comes amid rising cases of child self-harm linked to AI and AI-generated scams; the article does not disclose specific case counts or a timeline.

Anthropic blocked all foreign nationals from accessing Fable 5 and Mythos 5 after receiving a US government export control directive citing national security. The post doesn't specify which agency issued the order, whether existing enterprise deployments are affected, or if Anthropic will offer refunds or alternatives.

Anthropic shut off access to Fable 5 and Mythos 5 on Friday night, just days after launch. The US Commerce Department directed the shutdown, citing national security concerns over a potential Fable 5 jailbreak. The post doesn't disclose the jailbreak details or the directive text, but confirms the models are now offline.

The US government suspended foreign-national access to Fable 5 and Mythos 5 on national security grounds. Anthropic disabled both models immediately; other Claude models are unaffected. Anthropic calls it a misunderstanding and is working to restore access. DAIR.AI's Elvis Saravia said he dropped Claude entirely this week and stressed the importance of sovereign AI. The post doesn't disclose the specific export control clause or a timeline for restoration.

Anthropic posted a statement that the US government directed them to suspend access to Fable 5 and Mythos 5, covering both inside and outside the US, including foreign national employees. The post doesn't disclose the specific reason or directive details, only a link to the statement. Reddit comments compare it to ITAR-level export controls and note the irony given Anthropic's past calls for heavy regulation.

Anthropic immediately suspended all user access to Claude Fable 5, citing a US government directive. Other Claude models are unaffected. New chats default to the user's chosen model or Opus 4.8, and existing Fable 5 sessions throw errors. API calls also fail, with migration to other Claude models recommended. The post does not disclose the directive's specifics or a timeline for reinstatement.

Amazon shared its global data center water consumption for the first time: 2.5 billion gallons in 2025, at 0.12 liters per kilowatt-hour. That's a 2% drop from 2024 even as operations expanded. Amazon also claims better water efficiency than some Big Tech rivals, pointing to a chart in its report. The disclosure came right after Seattle enacted a data center moratorium that some Amazon employees had pushed for. The post doesn't name the rivals or show the full comparison data.

A group of independent musicians is suing Google, claiming it trained its Lyria music model on YouTube creators' uploads. Google's response is evasive: it says the terms of service allow it, but refuses to confirm or deny whether it actually did. The post doesn't disclose Lyria's training data sources or scale.

On June 10, China's MIIT issued an implementation opinion on 'AI + information communications.' It orders accelerating backbone network upgrades to 400Gbps/800Gbps between eastern, central, and western hub nodes, and deploying 400G metro networks with all-optical cross-connect to achieve millisecond-level latency for computing access. It also pushes AI-driven network tech: 5G-A/6G air interface intelligence, self-optimizing networks, space-based computing networks, and agent internet. Operators are encouraged to use AI for smart home assistants, 3D viewing, and other consumer services. The post does not disclose investment figures or a timeline.

The European Commission issued an interim measure on June 9, ordering Meta to give third-party general-purpose AI assistants free access to WhatsApp until its antitrust investigation concludes. Meta banned external AI from the WhatsApp for Business API in October 2025, then switched to a paid model in March 2026. The Commission sees the paid access as effectively continuing the ban and harming smaller competitors.

A German district court ruled Google is directly liable for AI Overviews content after one overview wrongly linked two publishers to fraud, and the cited linked sources did not contain the statements.

China plans to invest about 2 trillion yuan, or $295 billion, over five years to build nationwide data centers, with funding covering large-scale data center infrastructure for domestic AI development.

The Pentagon reinstated Alibaba, Baidu, and BYD as U.S. national security risks after removing the three companies in February; the RSS snippet does not disclose the legal basis, timetable, or company responses.

China’s MIIT and SASAC require key products including humanoid robots to complete application validation and enter routine deployment by the end of 2026, targeting more than 100 high-value scenarios and deployment at the ten-thousand-unit scale.

OpenAI said on Monday it has entered its third phase, naming three goals: automated AI researchers, faster economic growth, and personal AGI for everyone, while calling for an international body to manage AI risks.

A judge blocked Trump’s $100,000 H-1B visa fee, ruling that the six-figure charge was an unauthorized tax and ordering it suspended; the RSS snippet does not disclose the court, case name, or implementation timeline.

FT says AI can use CCTV data to identify targets, and Russia paused a surveillance system after the killing of Iran’s Supreme Leader; the RSS snippet does not disclose the system name, model mechanism, vendor, or timeline.

OpenAI outlined its third-phase plan with three goals: build an automated AI researcher, accelerate the economy, and give every person a personal AGI. Sam Altman and Jakub Pachocki said OpenAI internally believes AI systems may perform a significant fraction of its research by March 2028, while alignment, safety standards, and international coordination remain explicit conditions.

Bernie Sanders met with Sam Altman to discuss transferring 50% ownership of major U.S. AI companies to the public, while the article cites a Quinnipiac poll saying 80% of Americans are concerned about AI.

U.S. House lawmakers released a draft AI bill that would bar individual states from setting their own AI regulations and concentrate regulatory authority at the federal level; the RSS snippet does not disclose the sponsors, committee path, voting schedule, or the federal mechanism that would replace state-level rules.

Trump said he has discussed deals with AI companies that let Americans benefit from AI’s success, and CNBC reported the administration is discussing an OpenAI equity stake; the article does not disclose the stake size or deal terms.

The title says Trump signaled interest in the US owning stakes in top AI labs; the body is a Bloomberg 403 anti-bot page and does not disclose the target labs, stake size, policy mechanism, or timing.

Police in England and Wales were told to halt AI use in court statements until safeguards are in place; the RSS snippet cites the head of Police.AI but does not disclose the specific safeguards or enforcement mechanism.

Trump said the US may take equity stakes in AI companies, but the FT article body is a subscription page and does not disclose stake size, target companies, transaction terms, or policy mechanism.

The Intercept says the Pentagon runs an AI propaganda mill targeting Latin America; the snippet discloses the target region, AI-generated content distribution, and 100 Hacker News points, but not budget, vendors, model stack, or operational timeline.

New York State lawmakers passed a one-year moratorium on new large data centers, and if Governor Kathy Hochul signs it, the ban will cover projects with peak demand of at least 20 megawatts.

The title says the US National Security Agency is using Anthropic’s Mythos for cyber attacks; the RSS snippet only says Anthropic is in a legal battle with the Pentagon over the Claude model and does not disclose deployment scope.

Dario Amodei, Sam Altman, and Mustafa Suleyman signed an open letter urging Congress to require synthetic DNA and RNA sellers to screen orders for risky sequences; the RSS snippet does not disclose bill text, enforcement timelines, or screening thresholds.

MIT and USC researchers examined 4.5 million federal civil cases from 2005 to 2026, finding self-represented lawsuits rose from 11% in 2022 to 16.8% in 2025, while AI-text detector flags in sampled filings increased from 1% in 2023 to 18% in 2026.

France’s €110bn AI investment wave is driving a major data centre build-out, while investors warn that approvals and local opposition can slow the construction schedule.

A UN University report projects global data center power use will rise to 945 TWh by 2030, with AI accounting for 40%, while water use will increase to 9.3 trillion liters.

SpaceX won a tax exemption for Elon Musk’s $55bn Terafab AI chip plant; the post says residents in a Texas county oppose the project and threaten legal action, but does not disclose the exemption size or legal timeline.

Bernie Sanders argues in a June 1, 2026 op-ed that the public should own 50% of big AI companies; the post does not disclose a specific legislative mechanism or which companies would be covered.