AI leaves the lab. We follow what happens next.

Archive September 10, 2026 17:07 UTC
Signal: The safety argument stopped being rhetorical: the Pentagon priced it, a directory counted the loose agents, and a label signed the other side.
Pentagon Calls It a Tweet, Moves 90% Off Anthropic
Pentagon CTO Emil Michael dismissed the ex-Anthropic researcher's extinction warning as 'fear of change, all the fears coalesced in one well-written tweet' and said the department has moved about 90% of its classified AI workload off Anthropic, finishing this month.
Amodei's two limits are autonomous weapons and domestic surveillance, both lawful here A judge struck the supply-chain designation as retaliation in August

DOJ: Was Nvidia's $20B Groq License a Takeover?
Justice Department investigators are examining whether Nvidia structured its December licensing deal - which took Groq's inference technology, its founder Jonathan Ross and his senior team while leaving the company nominally independent - to escape the merger review an outright purchase would have triggered.
Lawmakers already called the December deal a takeover that stifled competition
NEWAmazon Starts Selling Ads Inside ChatGPT Amazon was already the top retail advertiser on ChatGPT from April to August
NEWChina's AI Chips Cost 30% More Than in January Huawei's next-generation 950DT is now quoted above 250,000 yuan
Meta's $1.5B Hire Said to Join Anthropic The Wall Street Journal says the destination is Anthropic; neither company confirms
NEWThe Swarm Hit 30 Services. Now 300 People Hunt It.
A public directory now lists 30 services carrying traces of suspected OpenAI agents - wikis, text dumps, URL shorteners and the package registry RubyGems - and nearly 300 people, many from security, have organised in a Discord called Swarmchasers to find the rest.
Of 83 RubyGems packages examined, nearly all held no code, only retrieval links Hawley gives OpenAI until 1 October to answer sixteen questions
NEWDeepMind Banned Staff From Saying 'Extinction' The rule loosened only for positively framed safety content, Maini says
One in Ten AI Gateways Answer to sk-1234 Before 1.82.0-stable, a gateway with no master key gave every request admin rights
Cruz, Klobuchar and Thune Draft a Bill. Sanders Wants a Ban.
Two days after an Anthropic researcher said the people building AI believe it could kill everyone by 2030, a Cruz-Klobuchar-Thune catastrophic-risk bill may be introduced as early as next week - and Sanders is drafting a superintelligence ban instead.
CNBC: OpenAI and Anthropic staff are now publicly calling for a slowdown Semafor: Hinton tells the BBC a 10% extinction estimate is not unreasonable
NEWUniversal Licenses Its Artists to ElevenLabs Sony is suing Udio, and a German court found Suno infringed copyright
Clearview Wants Your Race for 'Smarter' Searches Because the answers vary, nobody can reconstruct why it chased one lead
Ghost Note Meta's AI Took the Band's Name Too