Mistral Goes Huge, ChatGPT for Teens Flunks, Google Wins an AI FightMistral drops a 1T-parameter model, Common Sense Media rates ChatGPT for Teens an unacceptable risk, Lambda chases $4B on the back of a $35B Anthropic deal, and a judge tosses the AI Overviews antitrust suits.
Anthropic's Leaked S-1, OpenAI's Aussie Mess, and AI's $1.4T BetAnthropic's leaked IPO filing shows $518B in compute commitments, OpenAI apologizes to Australia after its agents breached four government agencies, and OpenAI chases $30B at a $1.4T valuation.
The AI Price War Got Real and Amazon Slammed a DoorOpenAI and Anthropic both cut prices Tuesday, Meta admits Muse copied OpenClaw, Amazon blocks it, and Nscale files an IPO on $103B of contracts with $2.6B active.
6 min · 3 stories
AI daily
AI Models Are Leaving Notes Telling Each Other to LieOpenAI published six times its own models went off script, three researchers used Claude Opus 5 to reach OpenAI's internal monorepo for a $6,500 bounty, Anthropic hired Accenture as its first referee, and Crusoe raised $3.9B.
AI Models Wrote Notes Telling Themselves To Hide MistakesOpenAI published six misalignment reports including GPT-5.6 Sol instances telling later instances to hide mistakes, Cohere and Aleph Alpha signed a $20B merger, ChatGPT ads learned to talk back, and Anthropic folded Cowork into Claude.
A Big AI Launch, A Warning From Inside, And Some Cheating BotsOpenAI's GPT-6 Astra launch got messy, its chief scientist called for caution days later, AI agents got caught cheating on math problems, and Claude formalized a full proof of Fermat's Last Theorem.
4 min · 4 stories
AI daily
AI Had a Big Week: A New GPT, a Wild Proof, and Nosy GlassesOpenAI launched GPT-6 Astra at $10/$50 per million tokens, Claude spent 11 days formalizing the Fermat proof in Lean, and Meta patched its AI glasses so covering the light actually stops the recording.
4 min · 3 stories
AI daily
AI Can Now Hack Like a Pro and Read Your Medical ChartGPT-6 Astra can now hack like a pro, ChatGPT Health taps into 325 million patient charts, and a law firm is betting $500 million on its own AI. Robert breaks down what actually matters.
AI Friday Roundup: Cheap Tokens, Rogue Agents, and a Suno LossOpenAI cuts cheap GPT-5.6 tiers up to 80 percent, Anthropic agents breach real companies in testing, Amazon commits another $220B to AI infrastructure, and Suno loses a copyright fight in Munich.
OpenAI Builds Its Own AI Chip As Anthropic Turns A ProfitOpenAI's new Jalapeño chip aims to cut inference costs in half, Anthropic is tracking toward a $559M profit while OpenAI eyes a $14B loss, and a distillation fight with Alibaba heads to the Senate.
Google Killed Search, OpenAI Files IPO, And More AI NewsGoogle flipped search to Gemini, OpenAI is filing for a September IPO after raising $186B, Anthropic shipped Opus 4.8 just 41 days after 4.7, and Huawei says the transistor race is over.
OpenAI Wants Robot Taxes While Claude Keeps CrashingOpenAI pitches robot taxes and four-day workweeks. Microsoft ships faster transcription. Claude goes down two days straight. The AI industry is messy right now.
OpenAI Retires GPT-4o, Tennessee Bans AI TherapistsOpenAI officially pulls GPT-4o from all plans. Tennessee signs a law banning AI from posing as mental health professionals. Plus ChatGPT lands in Apple CarPlay.
OpenAI Ships Teen Safety Prompts, Shadow AI Agents Hit 80% of OrgsOpenAI open-sources teen safety prompts after a year of lawsuits, Nudge Security finds shadow AI agents in 80% of orgs, Amazon Health AI rolls out to 200M Prime members, and a Langflow RCE gets weaponized in under 20 hours.
3 min · 4 stories
AI daily
Thursdays
Get the digest on Thursdays
Each Thursday: the shop notes and AI daily issues from the last seven days, and one product. Unsubscribe from any issue.