Trump delayed signing an AI security executive order that would have required government security reviews before model release; the post does not disclose review criteria, model scope, or a new signing timeline.

Samsung reached a last-minute agreement to avert a work stoppage tied to AI-related gains; the RSS snippet says the strike threatened Korea’s economy and the global AI boom, but the post does not disclose deal terms, amounts, or worker counts.

Box Elder County commissioners approved the Stratos Project, a 40,000-acre data center in Utah’s Hansel Valley projected to consume 9GW of power, while the post cites expert warnings, public backlash, environmental risks, and added strain on already taxed water supplies.

China banned Nvidia’s gaming chip during Jensen Huang’s visit, and the RSS snippet says Beijing aims to support domestic players including Huawei and Cambricon as they catch up to US rivals, but the post does not disclose the chip model or the scope of the ban.

The European Commission published draft guidelines on May 19 for classifying high-risk AI systems under Article 6 of the EU AI Act, using intended purpose as the main test and allowing exemptions for auxiliary tasks; the snippet gives the consultation deadline as “206月23日,” so the exact date is not disclosed cleanly.

Chinese courts have publicly backed workers in AI-replacement dismissal cases for a third time; in the Hangzhou case, an employee’s monthly salary was cut from 25,000 yuan to 15,000 yuan before dismissal, and the court found the employer failed to provide a reasonable placement and had illegally terminated the worker.

US regulators paused some cyber-related examinations of the largest banks after Anthropic’s Mythos model exposed new risks. The RSS snippet does not disclose the exam scope, delay duration, affected banks, or Mythos technical details.

A federal jury in Oakland ended Elon Musk’s three-week case against OpenAI and Sam Altman by ruling that the statute of limitations had expired, while Musk said he would appeal to the Ninth Circuit.

An Anthropic co-founder will release the first AI encyclical with Pope Leo XIV in May 2026, and the post says it focuses on AI technology and ethics, with 104 points on Hacker News.

Nvidia CEO Jensen Huang said Chinese authorities will eventually allow imports of US AI chips; the RSS snippet only says he made the comment days after joining President Donald Trump’s summit in China and does not disclose a timetable.

A jury rejected Elon Musk’s claims against Sam Altman and OpenAI over its shift toward a for-profit structure, finding he waited too long to sue; the post does not disclose the court venue, requested remedies, or overhaul terms.

The jury reached a unanimous verdict after about two hours in Musk v. Altman, finding two of Musk’s three claims time-barred and the third failed after a related dismissal; the advisory jury’s verdict was not legally binding, but US District Judge Yvonne Gonzalez Rogers accepted the decision.

Nine California jurors unanimously ruled that Elon Musk filed his claims against Sam Altman and OpenAI too late, and the RSS snippet does not disclose the requested damages or detailed legal reasoning.

Thomas Dietterich, chair of arXiv's computer science section, announced a rule that gives all listed authors a one-year ban when a paper contains confirmed unchecked LLM-generated content, and requires post-ban submissions to pass peer review before upload.

Anthropic will brief members of the Financial Stability Board on capabilities of its new AI model; the title says Mythos exposed cyber flaws, but the RSS snippet does not disclose flaw details, model parameters, or the briefing schedule.

arXiv will ban authors for one year if they let AI do all the work on a paper; the article says first-time submitters already need endorsement from an established author, but the provided body does not disclose further enforcement details.

Google DeepMind 宣布与新加坡政府达成国家 AI 合作，在新加坡推出多项计划，聚焦医疗健康、科学发现与教育。合作内容包括探索 AI 辅助临床医生、用 AlphaFold 和 Google Earth 推进东南亚传染病研究、为盲人及低视力跑者开发基于 Gemma 的跑步助手，并向中小学至初级学院教育者提供 Gemini for Education。

Musk asked the court to unwind OpenAI’s 2025 restructuring and sought up to $134 billion in damages; the jury will begin deliberating Monday, but its advisory verdict will not bind the judge.

ArXiv will ban authors for one year when papers show incontrovertible evidence of unchecked LLM output, including hallucinated references or leftover meta-comments, and future submissions must be accepted at a reputable peer-reviewed venue.

Meta will receive $3.3 billion in tax breaks for its $10 billion Louisiana data center; the post does not disclose the incentive mechanism, construction timeline, or compute use case.

Google updated its Search spam policy to classify attempts to manipulate generative AI responses in AI Overview or AI Mode as spam, and the RSS snippet names biased best-of listicles and recommendation poisoning as tactics while not disclosing the full enforcement details.

The UK Treasury, Bank of England, and Financial Conduct Authority warned that the most advanced AI models can run cyberattacks faster, at broader scope, and lower cost than ordinary professionals; the snippet says Bank of England Governor Andrew Bailey named Anthropic’s Mythos, but does not disclose test methods or quantitative benchmarks.

Anthropic is in a legal dispute with the US government over whether federal agencies will ban its AI models, and Bloomberg’s RSS snippet says the dispute has become a financial threat to Figma and other businesses.

Anthropic outlines two 2028 scenarios for US-China AI competition: if the US and allies expand their compute-chip advantage through export controls, theft prevention, and faster AI adoption, democratic states can maintain a 12-to-24-month technical lead.

Gallup found that more than 70% of Americans oppose AI data center construction in their area, based on a March 2026 survey of 1,000 U.S. adults; only 7% said they strongly favor new data centers.

Gallup found that 70% of Americans oppose data centers near residential areas, and 69 U.S. jurisdictions have issued moratoriums as projects face protests, regulatory disputes, higher wholesale power prices, and concerns over air pollution and water use.

A federal court in Southern California accepted a class action against OpenAI, with plaintiffs alleging that the ChatGPT website used Facebook Pixel to send query topics and cookies containing a Facebook unique ID to Meta in real time.

Chinese think-tank representatives asked Anthropic in Singapore last month to give Beijing access to Mythos, and Anthropic refused; the company has limited the vulnerability-finding model to the U.S. government and more than 40 organizations.

Trump called Jensen Huang on Tuesday morning to invite him to join the China trip; the White House’s Monday list of 16 CEOs did not include him, while Nvidia is still seeking approval to sell AI chips to China.

Medicare’s ACCESS creates a payment mechanism for an AI agent that monitors patients between visits, calls to check in, coordinates housing referrals, and confirms medication pickup. The RSS snippet does not disclose payment rates, rollout timing, participating states, or operational requirements.

Altman testified in the legal battle with Musk that Musk made “hair-raising” demands for control of OpenAI; the RSS snippet only says the startup’s CEO took the stand and does not disclose the specific control terms.

Attorneys general from six U.S. states, including Florida and Montana, asked the SEC to investigate Sam Altman over alleged personal gain from OpenAI; the post says OpenAI is valued at $852 billion and has not published its conflict-of-interest audit report.

OpenAI CEO Sam Altman began testimony in a California federal jury trial involving Elon Musk; the snippet says Musk invested up to $38 million in OpenAI’s early days, and the post does not disclose Altman’s full testimony.

Hangzhou Intermediate People’s Court ruled in China’s first unfair competition case over AI-written product-seeding notes, ordering Companies B and C to pay the platform RMB 100,000 and applying a four-factor test for generative AI service providers’ duty of care.

A senior South Korean policymaker proposed paying citizens a dividend funded by taxes on AI profits; the RSS snippet does not disclose the tax rate, payout size, legislative path, or implementation timeline.

China’s CAC, NDRC, and MIIT issued an implementation document on AI agent standardization, targeting privacy leakage, unauthorized actions, and loss of behavioral control from high-autonomy, high-permission agents, while tying the work to a 2027 target for new intelligent terminals and AI agent adoption above 70%.

Import AI 456 covers radical optionality for AI regulation and a Neural Computer paper, listing seven proposed governance tool categories, including transparency, reporting, audits, whistleblower protections, evaluations, model-weight security, and talent, while also noting Meta and KAIST prototypes using Wan 2.1 for CLI and GUI neural-computer experiments; the RSS snippet is truncated before full prototype results.

NHS will grant Palantir consultancy contractors working on the Federated Data Platform access to patient data; the RSS snippet does not disclose the access scope, duration, safeguards, or audit mechanism.

MIIT launched a pilot plan for AI ethics review and services, assigning four tasks and planning a national ethics risk monitoring service network for review practice, standards work, and multilevel governance.

Greg Brockman testified that Elon Musk pushed OpenAI in 2017 to create a for-profit arm and sought majority equity, board control, and the CEO role; Musk now asks the court to remove Sam Altman and Brockman, unwind OpenAI’s restructuring, and award up to $134 billion from OpenAI and Microsoft.