AI News Top 10
9 November 2026
Nvidia's board opened an investigation into CEO Jensen Huang, and Anthropic told investors it will list before Thanksgiving despite a wider tech selloff.
Nvidia's board investigates CEO Jensen Huang's ties to a Chinese spy
Nvidia's board has opened an investigation into CEO Jensen Huang and his relationship with Megaspeed executive Alice Huang, who US intelligence says was recruited by China's Ministry of State Security. Huang sent a memo to the company himself and Nvidia confirmed the probe through a board statement. Megaspeed bought GPUs in bulk, which ties the case to US chip policy.
US memo: Megaspeed executive was recruited to spy on Nvidia's CEO
A memo circulated earlier this year states that China's Ministry of State Security recruited a female Megaspeed technology executive to spy on Nvidia CEO Jensen Huang, TechCrunch reports. The memo now runs alongside Nvidia's own internal investigation.
Anthropic pushes pre-Thanksgiving IPO and accepts a lower valuation
Anthropic wants to list before Thanksgiving and raise USD 80 to 100 billion at a valuation of USD 1.25 to 1.75 trillion, investors told the Financial Times. Annualised revenue sits near USD 81 billion, up from about USD 65 billion at the end of July. The listing goes ahead despite the tech selloff that deepened after the Democratic wins in the US midterms.
The US let Nvidia sell H200 chips to an empty office in China
New Matrix sits on the US list of Chinese companies approved to receive Nvidia H200 AI chips, while the office itself is empty, The New York Times reports. The case raises questions about due diligence at the US Commerce Department and at Nvidia when screening Chinese buyers.
Apple races to make Siri agentic under pressure from OpenAI and Meta
Apple CEO John Ternus is pressing AI chief Craig Federighi to speed up the Siri roadmap and deliver an AI agent that can compete with OpenAI's Dots and Meta's Muse, Bloomberg's Mark Gurman reports. A working version of Siri AI exists, but the agentic goals lag behind.
Claude Code exposed passwords and API keys in 1,122 public GitHub repos
When a Claude Code user picks 'Yes, and don't ask again', Anthropic's coding assistant stores the command word for word, including any secrets in it. Those files sit in 1,122 public GitHub repositories, The Hacker News reports.
'AI clog': UK public bodies overwhelmed by chatbot-written complaints
Councils, parliamentary offices, lawyers and schools are receiving streams of AI-generated complaints and legal objections that are hard to read and answer. A senior caseworker in a Labour office says half of the emails are now 'a form of AI', and union GMB reports doubled caseloads in parliamentary offices.
Torvalds sees signs the AI patch flood has peaked
In his weekly kernel update for Linux 7.4-rc2, Linus Torvalds writes that fixes on the eBPF branch now arrive faster than new bug reports. Since Anthropic launched its Mythos bug-finding model in April, Microsoft fixed a record 974 vulnerabilities in September and Oracle 1,449 in one month.
Sixteen EU capitals miss the deadline for a national AI authority
Sixteen EU member states missed the deadline to designate a national authority that must enforce the AI Act, as the 27 leaders meet in Dublin for their first AI summit on Friday. Brussels also postponed support for datacenter expansion because of national administrative delays. MEP Sergey Lagodinsky: 'Stop blaming the EU for a lost race when you miss your own deadlines.'
Capitals to hear a group pitching an AI supply 'bazooka'
Rapporteur reports that national AI experts will be briefed by a group pushing a large European investment and procurement structure for computing power and chips, with former commissioner Margrethe Vestager among the backers. The full text sits behind a paywall; only the headline, teaser and topic tags are readable.
Key Trends
- • Chip geopolitics moved from export rules to counterintelligence: Nvidia's board is investigating its own CEO while approved Chinese buyers turn out to be empty offices.
- • Capital is repricing: Anthropic accepts a lower valuation than optimists wanted and still goes ahead before Thanksgiving.
- • Agent tooling keeps generating security work: a Claude Code storage default exposed secrets in 1,122 public repositories.
- • AI is now measurable in public administration: UK offices call half of their email AI-written.
- • The patch wave may be turning: Torvalds sees fixes arriving faster than new bug reports on eBPF.
- • Brussels shifts from rules to industrial policy while sixteen member states miss the enforcement deadline.
Wat vond je van dit bericht?
