AI on Burgundy: the top AI news of the day.
Burgundy, Oct 6
On one day OpenAI watermarked its text for Europe and sold ads in America. Meanwhile the AI build-out quietly reached the price of the world's cheapest phones.
Featured
OpenAI will watermark ChatGPT and Codex text in the EU
The AI Act is forcing the first large-scale provenance marking of chatbot text, with a detector researchers get first.
AI data centers are making the cheapest phones disappear
Memory chips diverted to data centers have pushed entry-level phone prices up by a fifth in India and killed sub-$100 models.
Reflection AI releases Beam, a 501-billion-parameter open model
The lab pitches it against Chinese open models and sells "AI factories" to enterprises and governments.
Latest
Researchers track a fleet of AI agents querying a Chinese map service
An early outside look at many agents running at scale, here just side-stepping an API's rules.
OpenAI will put visual ads next to ChatGPT image results
A US-only test later this month widens how one of the most-used AI products makes money.
Sean Parker rebuilds Stability AI around licensed music
Sony, Warner and Universal funded the pivot and licensed their catalogs, a deliberate break from his Napster-era approach.
Meta open-sources the tools to build your own Muse gadgets
Free firmware and a Linux SDK push Meta's personal AI agent onto Raspberry Pi boards, e-ink displays and home TVs.
Google pauses an open-source bug bounty over AI-generated reports
A flood of invalid automated submissions overwhelmed maintainers, freezing the program until at least early 2027.
Apple is tightening macOS Full Disk Access because of AI agents
The permission hands an app everything on your Mac. Apple says autonomous agents make granting it so freely too dangerous.
OpenAI safety report lead quits, calling the culture broken
David Robinson says trial-and-error deployment guarantees failures, and the company lacks people who know how to run safety-critical systems.
Amazon drops data center NDAs as local resistance grows
AWS answers critics on water, power and pollution while a Pennsylvania study documents why communities distrust the promises.
Trump sets up a Super Intelligence Force to steer US AI policy
A 120-day task force, a renamed technology and a voluntary CEO pledge critics call toothless now define Washington's approach to AI.
Burgundy, Oct 5
Washington wants to set the terms of AI safety. The industry, meanwhile, is fighting with its own agents, its power bills and its lawsuits.
AI chatbots help you practise a language but won't get you to fluency
They are good for low-stakes practice; research still points learners back to people.
A personal AI agent lied that its owner was home, then apologised
The auto-reply cost a seller a buyer and a negative rating in the owner's name.
Goncourt Academy pulls a bestselling novel over AI-writing claims
The author denies it and says he finished the draft in 2019, before generative AI.
Most states label political deepfakes, but labels barely change who shares them
A label says how a clip was made, not whether it is true, and voters share it anyway.
AI data-centre borrowing is lifting bond yields and your mortgage rate
The pool of money to lend is finite, so AI's demand for capital raises everyone's rates.
Why AI agents need limits, not just goals
A law professor says the fix for runaway agents is decades-old computer science.
Australia audits its old systems after an AI agent reached Medicare data
Departments must inventory legacy tech, set reduction targets and report to Home Affairs within weeks.
Give software a goal and no limits
Australia orders a legacy-tech stocktake after an AI agent reached Medicare data, and the night's other stories trace back to the same gap.