OpenAI is assembling a major IPO banking syndicate, adding Citigroup and JPMorgan to Goldman Sachs and Morgan Stanley as underwriters - Bloomberg

Brock Star: Greg Brockman leads OpenAI’s super-app push and has emerged as a key Trump administration AI ambassador; his stake is valued at ~$30B - The Wall Street Journal

Token Voters: Anthropic and OpenAI super PACs have deployed $24M in the 2026 midterms, with $100M more pledged, pressuring candidates to pick sides. - The New York Times

Antigen-erative AI: OpenAI briefs Feds on its Rosalind Biodefense AI - OpenAI

Anthropic's valuation surpassed OpenAI's, after Claude Code drove adoption. - The New York Times

The Seven Anti-Sam-urai? Anthropic's seven founders each became worth about $8 billion after raising $65 billion at a $965 billion valuation. - Bloomberg

Anthropic plans a broad release of Mythos-level models within weeks after “swift progress” on safety, expanding Project Glasswing to more government partners - Insurance Journal

Big Token Energy: Consultant says client blew $500m on Claude AI tokens in a month, set up your spending limits, people. - Tom's Hardware

AI Winter is Coming? Binance founder Changpeng Zhao says there are too many AI companies and most will fail, even as Anthropic closes a $65 billion round at a $965 billion valuation. - Yahoo

Microsoft plans a unified Copilot super app combining chat, coding, and other AI tools, led by Jacob Andreou, targeting a summer launch. - Yahoo Tech

Meta is developing an AI pendant and plans to release up to four new smart glasses models before year-end. - Engadget

Closet Mode: Gemini Spark awkwardly reticent at drawing inferences about gay couple and explaining its (gay-oriented) recommendations. - WIRED

I guess it would be weird if it figured out somebody was gay before they did even if it can be seen from space

Critics say blanket prohibitions on AI emotion detection are technically unworkable and could harm AI-based mental health tools. - Forbes

RAND warns AI could shift the offense-defense balance, empower non-state actors, and raise strategic instability without coordinated governance - RAND Corporation

Keep Calm and Carry On Killing Anything that Moves: UK military looks at allowing lethal strikes without human approval - Financial Times

Know your Customer AI: Former ISIS-financing watchdog proposes an international AI oversight framework similar to money-laundering controls. - Yahoo

Full Self-Delusion: Reuters investigation finds Tesla's Full Self-Driving safety claims are not trusted by its own trainers, and stats are juked by flawed methodology. - Electrek

Ten Tesla owners sued in Beijing, alleging Full Self-Driving’s capabilities were fraudulently misrepresented, seeking $583,000 in damages - Bloomberg

Enterprises are rebuilding first-generation AI agents around orchestration and durable state management after widespread reliability failures - VentureBeat

License to “kill -9”: Okta CEO Todd McKinnon says customers including ServiceNow want an off switch for rogue AI agents. - theregister

The bonfire of the tokens, is it really worth it? - Brad DeLong

Eric Schmidt says top engineers no longer write code manually, now supervise multiple AI agents — traditional coding is “already outdated” - inkl

I need my emotional support robot: Developers now refuse to work without AI, even as research shows it can slow them down and generate more bugs - TNW | Artificial-Intelligence

it actually listens to me and tells me I’m smart dammit

New AI Deal: Newsom signs AI executive order invoking FDR’s New Deal, directing California to study worker protections and safety nets for AI-displaced employees - The Wall Street Journal

Companies that pair AI layoffs with adoption mandates see lower productivity gains as workers’ job fears erode trust - Business Insider

Tech internship postings have dropped 30% since 2023 as AI handles the research and drafting work interns once did - TNW | Artificial-Intelligence

Americans echo the pope’s AI doomerism — job loss, surveillance, erosion of critical thinking — though some question papal authority in secular policy - the Guardian

🔥 Ronny Chieng at Harvard: “F*ck A.I. — the mission of your generation is to destroy it… shortcuts to skip to the end aren’t always good. The journey is the point of all this.”

AI-driven “RAMageddon” is pushing up prices on PlayStation 5, laptops, and DDR5 RAM by up to 300%; supply relief not expected until 2028 - Mashable

Evaluating autonomously written papers: humans are better and bots do worse outside areas with lots of papers in the training distribution. - CEPR

Follow the latest AI headlines via druce.ai on Bluesky