Policy & law
AI laws, policy and court cases. What is proposed, what has passed and who is affected.
AI news, explained simply
Today
Bessent blames Lutnick over Megaspeed export blacklist
Treasury Secretary Scott Bessent says his department has pushed for months to blacklist AI firm Megaspeed, but Commerce Secretary Howard Lutnick controls the decision.
Howard Lutnick’s ties to Nvidia draw ethics questions
Congressional members and officials are probing whether the Commerce secretary’s former firm, now run by his sons, helped Nvidia despite export‑restriction warnings.
US unions win new AI guardrails as agencies adopt generative tools
A survey finds almost 70% of U.S. government agencies use generative AI, while unions secure protections like retraining, oversight and veto power.
New private dense retrieval method adds 0.38 s to RAG pipelines
Researchers show a hashing‑based filter can keep queries private while only adding 6.5% latency to large‑scale retrieval‑augmented generation.
Yesterday
Democrats open probe into Nvidia’s China chip sales and Trump ties
House and Senate Democrats asked the Commerce Department to keep records on a deal that let Nvidia ship chips to China after a Trump‑approved tariff scheme.
AWS Quick adds custom permissions to limit data source types
Admins can now block or allow specific connectors in Quick, tightening governance for self‑service analytics.
Congress moves to tighten chip exports after Nvidia scandal
Republican and Democratic leaders back new bills that could restrict advanced AI chip sales to China before the end of 2026.
Nvidia hires prosecutor Joon Kim and law firm to probe former CEO
Nvidia retained Joon Kim and Wachtell, Lipton to investigate allegations about Jensen Huang after his resignation on Nov. 12, 2026.
Parents protest Google AI rollout in schools after lawsuit
Families in seven U.S. cities rallied outside Google’s Chicago office, demanding a halt to Gemini on school Chromebooks.
Google DeepMind staff ask UK government to trigger union ballot
UK DeepMind employees and two unions have applied to the Central Arbitration Committee to confirm support for a union vote covering about 1,000 staff.
DOJ opens antitrust probe into Nvidia after CEO resignation
The Justice Department is examining Nvidia’s market power while unrelated investigations target companies tied to a personal scandal involving Jensen Huang.
Nvidia founder Jensen Huang to testify before Congress
The former Nvidia CEO will appear at a closed‑door hearing by Thanksgiving after lawmakers asked him to answer questions about export controls and a recent scandal.
Sen. Tom Cotton asks US to block Chinese memory maker CXMT
Cotton wrote to Commerce Secretary Howard Lutnick urging placement of CXMT on the Entity List, citing national security and market concerns.
Australia plans law to force top AI firms to allow independent safety checks
The Albanese government will require frontier AI models to be audited, report incidents and publish safety plans under a new oversight regime.
Nvidia scandal fuels push for tighter US chip curbs before Trump‑Xi summit
Congress seeks stricter export rules on advanced AI chips after allegations that Nvidia CEO helped ship $7.5 billion to China.
Oossa · Newsletter
The week in AI, explained
Every Monday: the stories worth knowing, in plain language. Free, no spam.
Wednesday, November 11
Chinese university gets Nvidia chips despite US blacklist
A university linked to China’s military obtained Nvidia’s top GPUs last month, highlighting gaps in US export controls.
United blocks AI assistants from booking flights
United says personal AI agents can look up flights but cannot complete purchases, citing privacy and regulatory worries.
TechCrunch podcast says most announced AI data centers won’t be built
Brent Nelson of Ascend Analytics warns that the data‑center pipeline is wildly inflated and that permitting and power issues will keep many projects from materializing.
US data brokers sell $500 million of AI training environments to Chinese labs
Investigation finds at least nine American firms providing reinforcement‑learning worlds to Chinese AI companies, helping them cut compute costs.
House leaders seek Nvidia CEO briefing, warn of possible subpoena
Three House committee chairs want Jensen Huang to brief them on November 18 about security and his reported interactions with a Chinese intelligence-linked executive.
Nvidia CEO’s China stance shifts after 2025 export curbs
Jensen Huang began praising China and criticizing US chip limits after Washington barred Nvidia’s China‑specific chip in early 2025.
Sen. Bennet seeks update on Megaspeed investigation
The Colorado Democrat asked four federal offices for an update and urged them to consider penalties and stronger export-control enforcement.
U.S. and China set up AI risk dialogue, Bessent says
The two governments plan to share testing practices and alert each other to serious AI incidents, while continuing to compete in advanced AI.
Tuesday, November 10
New York opens AI safety registration ahead of 2027 law
The RAISE Act will require large AI developers to register, publish safety information and report serious incidents. New York opened its portal on November 10.
LexisNexis launches AI‑agent detection for government services
LexisNexis Risk Solutions adds 'Know Your Agent' intelligence to its ThreatMetrix and BehavioSec tools, letting agencies spot AI agents in digital interactions.
Monday, November 9
Trump says he learned of Jensen Huang’s alleged spy link only recently
At a Nov. 10 White House press briefing, former President Trump said he only heard about Nvidia CEO Jensen Huang’s connection to an alleged Chinese spy in the past few weeks.
California Sen. Scott Wiener elected to Congress, vows AI regulation
Wiener, who authored early AI safety bills in California, says he will push for federal rules to curb advanced AI risks.
Trump says DOJ warned AI exec to quit after alleged misuse
In a 60 Minutes interview, Trump claimed the Justice Department threatened a top AI company leader with jail, prompting changes at the firm.
Nvidia CEO Jensen Huang asked to brief Congress on Chinese spy ties
A House committee letter dated Nov 9, 2026 summons Jensen Huang for a closed‑door briefing on security concerns linked to a Megaspeed executive.
Muse has trained an AI model on users’ brain-wave data
Interaxon says it trained a model on about 50,000 hours of EEG recordings from its Muse headband. A Science investigation found privacy policies initially did not clearly explain this use.
Sunday, November 8
Nvidia CEO Jensen Huang admits affair amid spy allegations
Nvidia launches an independent board probe after Jensen Huang said he had an extramarital relationship with a Chinese executive suspected of espionage.
Elvis and Eminem estates sue AI music startup Suno
Rights holders claim Suno’s tool copies the singers’ voices without permission, citing a Tennessee law that bans voice cloning.
Friday, November 6
TechCrunch podcast examines election impact on AI industry
TechCrunch’s Equity podcast discusses how U.S. midterm results could affect data-center rules and scrutiny of Big Tech. Voters backed tighter restrictions in nearly all data-center ballot measures.
Meta pauses disputed publisher terms after backlash
Meta says it will consult publishers before changing its business terms, but revised wording still does not explicitly protect robots.txt preferences.
AI Firms Are Buying Companies’ Work Files to Train Models
Data brokers are seeking emails, documents and messages from businesses. Privacy and worker-consent questions remain, including in a disputed Spirit Airlines sale.
EU Commission says TikTok’s For You feed may breach DSA
The Commission says TikTok’s recommendations can create harmful “rabbit holes”. Its finding is preliminary and focuses on risk assessments and safeguards.
Thursday, November 5
Wednesday, November 4
US election results put data center projects at risk
Voters backed 38 of 44 ballot measures seeking data center restrictions, while new governors in key states have pledged tougher rules or project delays.
Pro-AI PAC Network’s Midterm Spending Had Limited Effect
Leading the Future backed candidates with millions of dollars, but its riskier bets mostly lost. More than $20 million remains in its bank account.
UK’s £400bn AI forecast came from Google-commissioned report
The government’s current estimate is £55bn to £140bn by 2030. The AI minister called even that range highly uncertain.