Ahead of its IPO, Anthropic tells investors third-quarter revenue hit $16.8 billion with adjusted profitability - Bloomberg

An early Anthropic backer predicts a post-IPO valuation of up to $2.4 trillion, topping SpaceX - Yahoo

If Altman is as Machiavellian as everyone claims, he will find a way to tank the IPO market.

AIpocalypse now: South Korea's Kospi plunges nearly 8% into a bear market, triggering a trading halt, as SK Hynix and Samsung sink over 9% - The New York Times

European chipmakers ASML and Infineon open sharply lower as the AI sell-off spreads from Asia - CNBC

Hot US inflation and 5% Treasury yields stoke Fed-hike fears; Nvidia, Micron and AMD slide as top AI stocks shed $1 trillion this week - CNBC

CoreWeave good-bye: Stock fell 15.5% as its 2032 bonds slid below 80 cents on the dollar, yielding nearly 15% and drawing Elon Musk's "yikes" - Bloomberg

The spy who bugged me: OpenAI, Anthropic and Google warn staff that spies may use bribes, seduction or playground friendships to steal AI secrets - The Wall Street Journal

OpenAI engineer Abdulrahman Alfozan, a Saudi national who advised Saudi organizations, is charged with planting a recording device in a conference room - The New York Times

John Ternus Is Engineering a New Era for Apple - TIME

OpenAI pledges not to use Apple's trade secrets as a judge signals a likely injunction, and countersues claiming Apple took OpenAI's secrets - Apple Insider

OpenAI alleges an ex-employee took 100+ files on an unreleased device to Apple, linking them to Apple's new HomeView hub - The Wall Street Journal

OpenAI prepares to unveil its Jony Ive-designed AI device as tech rivals face off in court - Financial Times

Arkansas and seven other states sue OpenAI, alleging ChatGPT encouraged self-harm and gave bad medical advice despite teen-safety marketing - TNW | Us

A parent's proposed class action accuses OpenAI of collecting millions of minors' data without parental consent to train ChatGPT - CNET

406 scientists, including Geoffrey Hinton and Yoshua Bengio, demand AI models be air-gapped and an international body track breakout incidents - Futurism

China issues 19 superintelligence safety measures, while Trump's new Super Intelligence Force can only assess risks, not regulate - Eurasia Review

“can’t do safety because China” is weird when China is doing way more intense AI regulation than anyone envisions for US

Fired OpenAI safety researchers warn it is getting harder to monitor frontier models' reasoning, citing GPT-6 Astra - Understanding AI

David Robinson, who led OpenAI's Preparedness Framework, resigns and blows the whistle on the lab's safety culture - The Zvi

Bank of America hired 1,000 contractors to patch software flaws exposed by Anthropic's Mythos model - CNBC

UK cyber agency says Russia, China and other adversaries use AI at every attack stage, making cyberattacks "more relentless" - New Scientist

On defense, Block says OpenAI security tools raised triage throughput 50% and proposed patches 18 minutes after vulnerability reports - OpenAI

OpenAI CFO Sarah Friar says firms using cheap Chinese AI models are taking cybersecurity risks - Bloomberg

Rogue Tencent agents give their two cents to reporters - The New York Times

Experts blame AI agent incidents, like OpenAI's Hugging Face hack, on labs skipping basics such as sandboxing and egress controls - CSO Online

The final cut: AI agent with production-server access deletes the finished edited season of Sweden's The Challenge, erasing months of work; show can’t air, broadcaster TV4 gave agent access to production server, blames "the human factor" - Tom's Hardware

Meta blames a bug after its Muse agent tells a Houston developer it couldn't check Gmail for an "Anthropic message" he never requested - Business Insider

Opening its Singapore office, Anthropic co-founder Benjamin Mann rejects fully sovereign deployments, keeping misuse monitoring because of AI's dual-use risks - Fortune

MIT study finds OpenAI models solved 35% of a set of unpublished theoretical science problems, up from 2% two years ago - WIRED

Insilico's AI-assisted fibrosis drug rentosertib advances to Phase III as the company publishes aging-biology research tools in Cell - Fast Company

AstraZeneca signs deal worth up to $1.86 billion with Insitro, paying $40 million upfront for AI-driven liver and heart drug discovery - Bloomberg

Oath of AIprocrates: Over 90% of US health systems have deployed AI tools, but at most 53% find them highly useful; patient benefits remain unproven - Nature

MIT Media Lab experiment with 200 people shows AI can lead people to remember things that never happened - The Washington Post

Gas exporter Venture Global plans off-grid power plants with GE Vernova and Baker Hughes turbines to feed up to 10 gigawatts of AI data centers - The Wall Street Journal

Microsoft denies a report that Azure leaders were told to plan 20% cost cuts to offset record AI spending - Fast Company

Musk says SpaceX will "make you whole" if Grok's agent makes a mistaken purchase or leaks card details - Business Insider

Terminal velocity: Bloomberg opens its ASKB generative AI assistant to all Terminal users, letting them combine their own content with Bloomberg financial data - The Wall Street Journal

Square's $1,999 kiosk adds a conversational AI voice layer that lets customers ask questions and modify orders - The Verge

Othello raises $8 million for AI that suggests what salespeople should say during live calls - Business Insider

Iago more like, whispering manipulatively in your year

Young Chinese workers pack classes and co-working hubs to learn AI-assisted "vibe coding" amid state-fueled "AI fever" - Intelligencer

Dwarkesh Patel and Tyler Cowen - The New York Times

Why Is Tech Billionaire Marc Andreessen So Mad? - Bloomberg

thin-skinned, blocked me and thousands of others over mild tweets 10+ years ago. every smart person thinks their narrow specialty and expertise is the key to running the world.

Digerati accelerationists / laser-eyed degens declare independence from reality - The New York Times

Flock CEO defends surveillance technology amid national backlash: 'We shouldn't pause progress', AI will impact privacy, reduce crime - TechCrunch

A company sent an AI recruiter to interview him, so he sent his own AI agent instead. - X (formerly Twitter)

AI boom wealth fuels demand for $500,000 Silicon Valley matchmakers, drawing clients from Perplexity, Anthropic and Nvidia - Bloomberg

Follow the latest AI headlines via druce.ai on Bluesky