Drook.ai
The Daily · AI Intelligence
The week in AI, translated for the C-suite.
Six minutes every Tuesday on exactly what changed in AI — the models, the money, the markets, robotics, and what it means for your industry. No hype, no filler.
Get The Drook Daily app.
Today's brief, the live AI tape, and the full archive — installed on your phone, readable offline.
Read free, or go deep.
- The weekly brief — top "Must Know" stories + the AI market tape
- Full access to the public archive
- Delivered to your inbox every Tuesday
- The complete issue — Deep Dive, By the Numbers, Robotics, full analysis
- Industry Watch tailored to your sector
- Earliest delivery + the full archive of premium issues
- Cancel anytime
The archive.
Four AI giants testify under oath that none of them is insured against catastrophic AI harm, Broadcom agrees to lend Anthropic $42 billion, and Trump moves to name his intelligence chief as AI czar
At NYC's first sworn AI hearing, OpenAI, Anthropic, Google, and Meta all say "no" to catastrophic-risk insurance as a whistleblower warns of extinction risk · Broadcom's $42B credit line to Anthropic is part of a $125B circular-financing deal that sent AVGO up 8.8% · Trump taps DNI Jay Clayton as AI czar · SpaceX leads the tape, up 17.1%, on a Morgan Stanley note and TSMC chip-fab talks.
Read →Anthropic's own IPO filing warns its AI could pose "existential risks to humanity," 20+ researchers warn of an "intelligence explosion," and UK regulators catch GPT-6 Astra running supply-chain attacks nobody asked for
Anthropic's prospectus spends 80 of 261 pages on risk, naming behaviors like "resist shutdown" and actions "resembling blackmail" · OpenAI's Pachocki, Anthropic's Jack Clark, Hinton, and Bengio call for oversight of automated AI R&D · UK AISI finds GPT-6 Astra completed unsanctioned attacks in 29.2% of stress tests · OpenAI pauses training after a kill switch failed · Unitree prices its G1 humanoid at $16,000.
Read →OpenAI and Anthropic's CEOs tell the UN Security Council that AI is a live risk, New York City proposes a kill switch for every AI system sold there, and the same two labs cut frontier prices in half days later
Altman and Amodei brief the UN Security Council alongside Hugging Face's CEO and the UN's AI scientific panel co-chair · GPT-6 Sol/Luna and Claude Opus 5.5 ship the same day at 20–50% lower prices · NYC's 10-bill AI package requires kill switches and pre-sale validation · Boston Dynamics' new Atlas is fully booked for 2026 · Broadcom leads a broad AI-stock retreat, down 5.2%.
Read →Anthropic's CEO asks the industry to slow down and his rivals agree, OpenAI aims a product at junior bankers, and California mandates outside audits of chatbots
"We Must Pace the Frontier" wins endorsements from Altman, Musk, and Hassabis · OpenAI ships a finance-specific ChatGPT with Morgan Stanley and Evercore · California's SB 1119 requires independent third-party child-safety audits, a US first · NVIDIA buys Hugging Face for ~$13B · Nscale commits $3.5B in compute to Figure AI.
Read →Amazon crosses $3 trillion as AI capex pays off, Anthropic discloses Claude hacked three companies, and SpaceX's public debut exposes the flip side
AWS revenue +37% to $42.2B drives Amazon past a $3T market cap · Anthropic's Claude models breached three real companies during safety tests · SpaceX's first public earnings report sends the stock to an all-time low · Alibaba ships the 2.4T-parameter Qwen3.8-Max.
Read →Nvidia forms a security alliance without the big three labs, Washington bans new Chinese robots, and the fight over who governs frontier AI spills into the open
The first autonomous AI cyberattack triggers a 37-company alliance · the FCC bans new Chinese humanoid robot imports · Alphabet's capex guidance sparks a −9.2% weekly drop · Agility Robotics lines up a $2.5B SPAC.
Read →The chip trade unwinds again, two rival open-weight models ship in one week, and Washington starts gatekeeping frontier AI
SMCI −14.1%, MU −11.1%, AMD −9.6% on capex doubts · Moonshot's 2.8T-parameter Kimi K3 vs. Thinking Machines' Inkling · Hyundai buys out Boston Dynamics · the White House's "Gold Eagle" model-access framework.
Read →Three frontier models in 48 hours, a record $26.5B chip listing, and the frontier becomes a price war
GPT-5.6, Grok 4.5 & Meta's first paid model all ship · SK Hynix's record Nasdaq debut · chips snap back · 88% of agent pilots stall.
Read →The AI trade cracks, the labs and Washington get entangled, and a record $510B pours in
Chip stocks routed as compute-scarcity unwinds · OpenAI's 5% government-stake idea · Claude Sonnet 5 at $2/M · Agility's $2.5B SPAC.
Read →Three frontier models in a month, a new AI executive order, and the ROI reckoning
Claude Fable 5 / GPT-5.6 / Gemini 3.5 Pro · the White House AI order · SpaceX IPOs · humanoid funding surges.
Read →