Austin walks through six stories. Executives from OpenAI, Anthropic, Google and Meta testified under oath before the New York City Council, and OpenAI started hiding a watermark in ChatGPT text for EU users. Utah is letting an AI app write acne prescriptions on its own after a supervised ramp. On the money side: former Groq engineers sued over Nvidia's $20 billion deal, chip startup Etched is reportedly fielding bids up to $50 billion, and Reflection AI unveiled Beam, a US open model aimed at China's best.
AI labs testify under oath before the NYC Council
Watch from 1:56
- What happened
- OpenAI, Anthropic, Google and Meta sent executives to testify under oath before the New York City Council, only after the council threatened subpoenas. All 51 members sat as a committee of the whole on Monday and asked about the chance of an AI catastrophe. OpenAI's representative said he didn't know and that no level of risk is acceptable; Speaker Julie Menon criticized that answer. Anthropic's Logan Graham and Meta's Shane Cahill gave no number, and Google's Alex Friend said no rigorous method exists yet. Former Google DeepMind researcher Alex Turner put the odds of an eventual AI takeover at roughly 1 in 3. On liability, OpenAI, Anthropic and Meta gave no direct answer; Google said if it's illegal without AI, it's illegal with AI. Menon demanded written follow-ups. Bills on the table include a whistleblower reward program, a right to sue over harm from AI agents, and independent testing.
- The claim
- The companies say they test their models for risks ranging from cyberattacks to loss of control.
- Austin's take
- Austin thinks asking for doomsday odds focuses on the wrong thing. He compares AI to the early internet, says the effort should go into skilling people up, and calls city-level restrictions on labs built elsewhere a bit absurd, warning they could mostly limit New Yorkers' own access.
- Why it matters
- Local governments are getting AI labs on the record under oath, following California's recent bills and subpoenas. Whether a city has real leverage over models built elsewhere is an open question.
- What to do
- If you're in New York City, watch the whistleblower, right to sue and independent testing bills, and look for the labs' written follow-ups.
- Not confirmed yet
- It's unclear whether New York City's jurisdiction can meaningfully affect labs that build their models elsewhere.
ChatGPT text gets a hidden watermark in the EU
Watch from 6:35
- What happened
- OpenAI said Monday that over the coming weeks, eligible ChatGPT and Codex output in the EU will carry TextGrain, a statistical signal hidden in word choices that readers can't see but a detector can. API customers worldwide can turn it on, but it's off by default. OpenAI plans to open source the technology, and at first only approved researchers and organizations get the detector. ChatGPT joins Anthropic's Claude and Google's Gemini in marking its own text. In OpenAI's own test at a 1% false positive rate, the detector found the mark in about 95% of 400 token passages and about 80% of 200 token passages. Swapping just 25% of the words for synonyms dropped detection to 17%.
- The claim
- OpenAI says this meets the EU AI Act's transparency rules for AI generated content, and that a missing watermark does not prove human authorship.
- Austin's take
- Austin says this makes little sense for text, since he finds AI writing easy to spot and tools like Pangram already exist. He sees watermarks mattering more for deepfake video, voice and images, and calls this an example of government moving slower than the market.
- Why it matters
- Teachers, employers and platforms want a way to tell AI writing from human writing, but the numbers show light editing largely defeats the mark.
- What to do
- Don't treat a missing watermark as proof a human wrote something. If you use the API, decide whether to switch it on.
- Not confirmed yet
- The accuracy figures come from OpenAI's own testing.
Utah lets an AI write acne prescriptions
Watch from 10:25
- What happened
- Nola Health said Monday its NolaDerm app can issue topical acne prescriptions to Utah adults under an agreement with the state's Office of Artificial Intelligence Policy, at a pilot price of $4.99 a month. Two physicians approve every prescription for at least the first 100 patients. Then the AI prescribes directly with weekly doctor review up to 500 patients, after which doctors audit at least 10% each month. No pills and no severe acne are covered. The waiver runs for one year. Nola's medical advisor has named flu and uncomplicated urinary tract infections as possible next steps.
- The claim
- The founder says a licensed clinician used to sign every prescription and the AI can now make that decision itself. Nola says the AI chooses among preapproved treatment protocols rather than making free form decisions.
- Austin's take
- Austin is excited, expects most cases are routine refills, and hopes it frees doctors for serious conditions. His concern is whether it raises prescription rates as it expands, and whether it lets financially motivated prescribers scale without doing the work. He thinks it's a net positive overall.
- Why it matters
- This is a template for AI making clinical decisions without a doctor signing each one. The real question is who signs off as the list of conditions grows.
- What to do
- If you're in Utah, know what the service covers and doesn't. Everyone else: watch whether flu or UTIs get added next.
Ex-Groq engineers sue over Nvidia's $20B deal
Watch from 15:35
- What happened
- Two former Groq engineers filed suit Monday in Delaware's Court of Chancery against Groq's board, naming founder Jonathan Ross and President Sunny Madra. In December 2025, Nvidia paid about $17 billion in licensing fees plus roughly $3 billion in stock bonuses to bring over about 200 Groq engineers, about $20 billion in all, in a deal described as non-exclusive. The suit says the board skipped a stockholder vote Delaware law requires and left Groq a hollowed-out shell, and that board conflicts left common stockholders billions worse off. The Justice Department is already looking at the deal, per a September New York Times report. Both plaintiffs left Groq before the deal but kept their stock.
- Austin's take
- Austin calls it a license and hire move used to buy startups without technically buying them. He suspects the plaintiffs may be opportunistic after missing a big payday, says litigation is expected with numbers this big, and wishes Groq and Nvidia more luck.
- Why it matters
- If a court says these deals need a shareholder vote, Big Tech's playbook for absorbing AI startups gets much harder, and common shareholders could gain leverage.
- What to do
- If you hold common stock or options at a startup, understand how a license and hire deal could affect you.
- Not confirmed yet
- These are allegations. Groq and Nvidia hadn't responded in the reports.
Etched reportedly fields bids up to $50B
Watch from 20:29
- What happened
- TechCrunch reported Monday, citing sources, that established investors have offered to value chip startup Etched at about $40 billion, with lesser known offers around $50 billion. Etched was valued at $10.3 billion in July after a $300 million round led by Sequoia and at $21 billion in September when Jane Street led $700 million. A report in July said it had $1 billion in customer orders, and Jane Street is both an investor and a customer. No round has closed.
- The claim
- Etched says its chips process tokens faster and cheaper than Nvidia's.
- Austin's take
- Austin is thrilled, says the jump shows how hard money is chasing Nvidia alternatives, and recommends the founders' Invest Like the Best interview. He wonders whether Nvidia, after spending $20 billion on Groq, could make a play for Etched.
- Why it matters
- Nearly a fivefold valuation jump in three months signals huge demand for alternatives to Nvidia for running AI.
- What to do
- Treat the figures as reported offers, not a done deal, and watch for a closed round or acquisition news.
- Not confirmed yet
- Based on unnamed sources. The report doesn't say who's bidding or whether the money goes to the company directly.
Reflection AI's Beam takes on China's open models
Watch from 26:10
- What happened
- Reflection AI unveiled Beam, a 501 billion parameter mixture of experts open model with 23 billion parameters active per token, trained on 23.8 trillion tokens. Early access opened Monday. Weights are due later in October under the Apache 2.0 license after red teaming. Reflection was recently valued at $25 billion.
- The claim
- Reflection says Beam beats GLM 5.2 on some tasks using a third to a quarter of the hardware, approaches Qwen 3.8 Max, still trails frontier models like Anthropic's Claude Fable 5.1, and is the first open model from a US startup to match the best Chinese models.
- Austin's take
- Austin calls it the strongest US bid yet in a space China has dominated. He argues open models level the playing field, worries about lock-in from Chinese open models, and notes a model like this could potentially run slowly on a MacBook.
- Why it matters
- Open weight AI has been China's game. A competitive US open model gives developers a non-Chinese option they can run themselves.
- What to do
- Wait for the weights and independent benchmarks before building on it.
- Not confirmed yet
- The performance claims are Reflection's own benchmarks, and you can't download the model yet. It matters only if the weights actually ship.
