Utah lets AI issue prescriptions without direct doctor oversight

Share
Utah lets AI issue prescriptions without direct doctor oversight

Utah signed the first US deal where an AI makes the prescribing decision itself, while a top robotics firm handed its keys to a frontier-AI veteran — and OpenAI started delivering on its 28-day shipping pledge.


Utah has approved the first US program where an AI makes the prescribing decision itself. The state's Office of AI Policy signed a regulatory-mitigation agreement with startup Nolla Health — not new legislation, but a waiver under Utah's 2024 AI Policy Act learning-lab sandbox — letting its app scan a patient's face, score the acne severity, and pick a treatment from eight approved topical medications without a physician signing off on each prescription. The scope is deliberately narrow: Utah adults 18 and older, mild-to-moderate acne, topical meds only (no pills, no isotretinoin), a 12-month pilot renewable twice. Oversight ratchets down in stages rather than disappearing — physicians approve the first 100 prescriptions, then review the next 400 weekly after the fact, then audit at least 10% monthly once the pilot runs fully automated; every prescription goes to the pharmacy under the overseeing physician's license, flagged as AI-generated, and advancement requires at least 95% agreement between AI and physician judgments with zero missed hard stops. Patients pay $4.99 a month for the app and $49.99 a month for compounded medication, against a $150 to $300 dermatologist visit. Why it matters: "AI replaces the doctor" just became a signed agreement with numbers attached, and the opposition is already on record — the AMA says autonomous systems "shouldn't be trying to practice medicine," and Utah's own medical licensing board warned a similar earlier pilot "places Utah citizens at risk." Whether the sandbox holds will be decided by the audit data, not the press release.


Boston Dynamics has named former Amazon AI executive Rohit Prasad as its CEO. Prasad spent 12 years at Amazon, most recently as SVP and head scientist for Alexa and artificial general intelligence, where he oversaw the Nova foundation models; he takes the helm October 7, succeeding Robert Playter, who ran the Hyundai-owned robotmaker for more than 30 years. The appointment is framed as a bet on physical AI, with Hyundai planning to put electric Atlas humanoids into its Georgia plants by 2028 at a target of up to 30,000 units a year. "By merging advanced AI and robotics, we will create meaningful value in the real world," Prasad said. Why it matters: the people who built the frontier-model world are now running the hardware robotics companies — and Boston Dynamics is doing it while still losing money on Spot and Stretch deployments, so the CEO seat goes to an AI strategist precisely when the humanoid transition needs to turn pilots into product.


OpenAI delivered day one of its 28-day shipping pledge: GPT-6 Astra and GPT-6.1 Sol now stream about 50% faster. OpenAI's Tibo Sottiaux announced the change as the first day of the quality-of-life sprint he promised earlier this month; the team "optimized inference," pushing streaming throughput to roughly 50 tokens per second from about 30, and the speedup applies by default across OpenAI products and Sign in with ChatGPT partners such as OpenCode, Pi, Amp, and Devin. The caveats matter: the numbers are OpenAI's own and haven't been independently timed, the change is subscription-only with the API unconfirmed, and faster streaming is a perceived-latency win, not a task-completion one. We covered the pledge when it landed — OpenAI's Codex lead pledges 28 days of daily fixes or full resets. Why it matters: after DevDay capacity meltdowns and global usage resets, responsiveness is the most user-visible quality lever OpenAI has — and day one shipped something users can feel within a single prompt.

What to watch: whether Utah's audit gates hold their 95% agreement bar once the pilot clears 100 prescriptions, and whether day two of OpenAI's clock brings another visible win.

Should a state sandbox let an AI prescribe before any malpractice case has tested it? Tell us in the comments.

Read more

Akhetonics says its all-optical CPU reaches a customer in 2026

Akhetonics says its all-optical CPU reaches a customer in 2026

Light-based computing keeps promising more than it delivers — but one Munich startup has just put a date on its bet, and the interview laying it out is doing the rounds on Hacker News this week. Akhetonics says it will deploy its first commercial machine with a major customer by the end of 2026, with several more planned for 2027. The company, founded by Michael Kissner and Leonardo Del Bino, is building a computer where data enters as light, is switched as light, and circulates through memory

The Week in AI — October 5–11, 2026

The Week in AI — October 5–11, 2026

Every big claim this week turned out to rest on fine print more interesting than the headline: revenue only the company reporting it can define, safety tests sandboxed while the product keeps the web, and a Pentagon phase-out nobody would confirm until reporters kept asking. The week's top 5 1. OpenAI's revenue was $20 billion below the numbers everyone quoted — and the gap was definitional. The Financial Times reported Thursday that OpenAI's annualized revenue runs roughly $20 billion unde

Drone strike shuts a third Yandex data center, taking YandexGPT offline

Drone strike shuts a third Yandex data center, taking YandexGPT offline

Russia's largest tech company is learning what the AI era's infrastructure war looks like from the receiving end — three data centers in four days, and with them much of the cloud layer Russian businesses run on. A Ukrainian drone strike knocked out Yandex's data center in Vladimir early Sunday morning, the third of the company's facilities hit since October 8. The site — reported at roughly 50 MW and designed for about 2,880 server racks — stopped operating completely after the attack, Yandex

Agent teams cost up to 5x more, barely score higher

Agent teams cost up to 5x more, barely score higher

The multi-agent hype train hit a benchmark this weekend — and the grid and the trucking regulators had quiet weeks of their own. Vals AI put agent teams head-to-head with single agents on its Vibe Code Bench, and the teams cost between 1.8 and 5.1 times more for almost no extra quality. The evals company ran GPT-6 Sol and Claude Opus 5.5 solo and in teams across 50 apps at two reasoning efforts; out of four comparisons, only one was statistically significant — Sol at medium effort, where the t