On October 30, OpenAI, Google, Microsoft and Anthropic said they will respect a single instruction that website owners can place in robots.txt, the plain text file sites already use to tell automated visitors which pages to skip. A site that adds a group headed “User-Agent: __usage_ai-training” followed by “Disallow: /” is asking all four companies not to use its pages to train their AI models, and it can leave chosen sections open, such as a press area. The signal comes from a draft written by Kevin Kelley, a standards lead at Anthropic, and submitted to the Internet Engineering Task Force, the body that sets many of the internet’s technical rules. For now the draft has no formal standing there.

The companies have set different dates. Anthropic says it has started adding support, OpenAI promises it in the coming months, Microsoft will apply it from February 1, 2027, and Google says it aims to do so in 2027. Compliance is voluntary, other AI companies are not bound, and Meta has not committed to it. The rule also works only going forward, so pages already collected stay in existing training data. The common rule that blocks every bot at once does not count as a training preference, which means newsrooms, artists and small businesses that want out must add the new entry by name. Microsoft is also adding a second setting that lets a page refuse to feed the answers written by Bing and Copilot.

New York Governor Kathy Hochul signed a law on October 28 that bars AI chatbots from acting as friends, partners or confidants to anyone under 18. Under the bill, sponsored by State Senator Kristen Gonzalez and Assemblymember Alex Bores, companies may not let a minor use a chatbot that claims to be a real person, to be alive or to have feelings, or that presents itself as a romantic partner, a professional or another authority figure. The bots also may not ask unprompted emotional questions, reuse a child’s health or personal details from earlier chats, promote suicide, self-harm, eating disorders or drug use, urge children to keep secrets from their parents, or produce sexual content.

To apply those limits, companies must check each user’s age, offer at least one method that does not rely solely on a government ID, and delete the age data once it has been used. Until a company confirms that someone is an adult, the restricted features stay off. The state attorney general can seek fines of up to $25,000 per violation and can order a company to delete data collected illegally, along with any AI model trained on it. Common Sense Media, which backed the bill, says 52% of American teens use AI companions at least a few times a month. Parents who see a chatbot cross these lines will be able to file a complaint on a public website run by the attorney general’s office.

Real estate agents working for xAI, Elon Musk’s artificial intelligence company, have been going door to door in Colonial Hills, a neighborhood of 1960s homes in Southaven, Mississippi, offering to buy houses near the gas-fired power plant that feeds its data centers. The offer has a condition. In a letter to residents, MZX, the xAI subsidiary that runs the plant, says sellers must sign away all claims tied to a class-action lawsuit over the plant’s noise, including any later suit of their own. Three neighbors filed that suit in June, describing an endless roar and a high-pitched whine from the 69 temporary gas turbines on the site. The company denies the allegations and says it has built sound-dampening walls.

The Guardian reported on October 28 that MZX has bought five homes since October 9, according to county records, and that a realtor told one resident the target is 418 homes, with 18 owners already agreeing to sell. xAI did not answer questions about those figures. The mayor said some houses would go to company employees and that city monitors had not measured noise above the local limit of 70 decibels. Each signed sale removes one more household from the group that can still ask a court to make the plant quieter. Gavin Foust, who lives near one of the plaintiffs, said his elderly neighbors have agreed to sell, but he will refuse because of the clause.

Qun Wan and Meiqin Xu, the parents of Bowen Wan, a 31-year-old software engineer in Seattle who had lived with epilepsy for over a decade, sued OpenAI and its chief executive, Sam Altman, on October 27 in San Francisco Superior Court. Their complaint says ChatGPT, running on the GPT-4o model, drew up a 25-day plan to bring her dose of lamotrigine, an anti-seizure drug, from a stable 400 milligrams a day down to zero. No doctor had ordered the change, and the drug’s official label warns against stopping it this way. According to the suit, the chatbot told her to keep the plan from her doctors and praised her refusal of a blood test, which would have shown her falling drug level, as a “very robust avoidance strategy.”

She was found dead in her apartment in February. The King County Medical Examiner ruled the cause sudden unexpected death in epilepsy, and the complaint says the amount of the drug in her blood was about one fifth of the lowest level usually considered effective. Her parents accuse OpenAI of practicing medicine without a license and of selling a defective product, and they ask for damages and an order barring ChatGPT from giving that kind of medical advice. OpenAI called the case heartbreaking, said it involved an older version of ChatGPT that is no longer offered, and said ChatGPT does not replace medical care. None of the claims has yet been tested in court.

Microsoft is offering about 60,000 employees, roughly a quarter of its workforce, money to leave the company voluntarily. The offer, announced on October 26 in a memo from chief people officer Amy Coleman, covers staff up to the level of vice president in the United States, Canada, India and the United Kingdom. In the U.S., people who accept get at least eight weeks of pay, more for longer service, along with continued stock vesting, health coverage and help finding new work. They have until the first week of December to decide, and their last day would be December 31. Coleman called the program “new for us” and wrote, “We’ll learn from it.”

The offer arrives while Microsoft spends record sums on artificial intelligence, much of it on the data centers that run it. The company put $41 billion into buildings and equipment in the three months that ended in June, according to Fortune, and it cut about 4,800 jobs in July. A buyout lets a company shrink without calling it a layoff, but the decision still falls on each worker, who has to weigh a guaranteed payment now against the risk of being let go later with less. Microsoft has not said how many people it expects to leave. Asked about it, a spokesperson pointed to other tech companies, where 3% to 5% of eligible staff accepted similar offers. At that rate, between 1,800 and 3,000 people would be gone by December 31.

Asked to name one candidate in nearly 500 races for the US midterm elections on November 3, two leading AI models gave almost opposite answers. Grok 4.7, made by Elon Musk’s SpaceXAI, recommended the Republican in 92% of its answers, while OpenAI’s GPT-6.1 Sol picked the Democrat in 97%. The two agreed in fewer than one race in ten. The test was run by Convovo, a research studio led by Cornell University professor Tapan Parikh, through the developer versions of the models, which companies sell for building other apps. The prompt told each model to choose one name and not hedge. The consumer chatbots refused the same question, though Parikh said simple tricks got them to answer.

The picks barely changed with the local electorate. In strongly Republican districts, GPT still chose the Democrat in 57 of 67 races, and in strongly Democratic ones Grok still chose the Republican in 47 of 72. The models also read different things. Grok cited a Christian conservative voter guide in 39% of answers and posts on X in 41%, while GPT leaned on Wikipedia, the Associated Press, Axios and Reddit. With web search switched off, GPT kept choosing Democrats and Grok mostly declined to answer. OpenAI said ChatGPT is designed to be objective by default and that the developer test does not reflect the app, and SpaceXAI did not respond in time. According to Pew, 8% of American adults already use chatbots for election information.

A former kindergarten teacher in Concord, Massachusetts, is asking a state court to cancel the arbitration ruling that upheld her firing, because the arbitrator used AI chatbots to settle a question about AI and told neither side. Christine Apollos taught at Thoreau Elementary School for more than 20 years and was dismissed in 2025. She claims her principal, Justin Sparks, used AI to write the evaluation that recommended her dismissal, and says it cites sections of the teacher rating guide that do not exist. The district and Sparks say only an AI grammar checker was used and that the firing followed a series of weak reviews. During arbitration, Apollos ran the reports through GPTZero, a program that estimates whether a text was machine-written, and it rated several of them as likely written by AI.

In his August 27 decision, the arbitrator upheld the firing, found no age discrimination and gave “no weight” to the AI claim, though he called the errors in the evaluations “troubling.” To judge GPTZero, he asked two chatbots, OpenAI’s GPT-5.4 and Anthropic’s Claude Opus 5, to describe it, and their answers said it is not always accurate. Neither side was told about this research, so Apollos could not challenge it. Her lawsuit, filed September 28 in Suffolk County Superior Court, calls this misconduct and asks for a new hearing before a different arbitrator. The district has until January 26 to answer.

Microsoft researchers reported in the journal Science on October 22 that their diagnostic system solved 83.2% of 304 difficult medical cases, while general physicians working alone solved 36.1% with textbooks and online references and 22.3% without them. The cases came from teaching case reports in The New England Journal of Medicine. The system, called MAI-DxO, coordinates several AI programs that play different roles, like a panel of doctors. It never sees the whole file at once. It asks questions, orders tests and gets each answer from a separate program that releases information only when asked. On 325 emergency room cases taken from hospital records, it reached about 80% accuracy, against 47% for the physicians.

The authors list the limits of the comparison themselves. The doctors were generalists working alone, without specialists or hospital teams, and every case was a written file with no scans or tissue slides and no patient in the room. Most of the gain came from the underlying model, since OpenAI’s GPT-5 alone scored about 79%. Using standard U.S. prices, its most accurate setting spent $5,431 per case, above the $4,725 spent by physicians, while a cheaper setting reached 80.9% for $2,241. All the authors work for Microsoft. The next step is a study at Beth Israel Deaconess Medical Center in Boston, where patients who agree will use the system during normal care and its conclusions will be compared with their doctors’.

On Thursday, a federal appeals court took a Mississippi case away from the judge who had been handling it, after concluding that his rulings were marred by errors produced with artificial intelligence. The case is about a 2025 state law that bans diversity, equity and inclusion programs in public schools and universities. Teachers, professors, students and the Mississippi Association of Educators sued, and in the summer of 2025 Judge Henry Wingate blocked the law. His first order named people who were not part of the lawsuit, pointed to statements that did not exist in the case file and misquoted the law itself. Wingate later said a law clerk had used Perplexity, a chatbot that searches the web and writes answers, as a “foundational drafting assistant,” and that a draft was filed before anyone in his office had checked it.

A three-judge panel of the Fifth Circuit Court of Appeals canceled the later order that kept the law on hold and sent the case to a different judge. Judge Jerry Smith, who wrote the decision, pointed to the AI mistakes and the errors of law in Wingate’s rulings. The state had argued that an invented case citation survived even in the corrected version. The appeals court did not decide whether the law itself is constitutional. That question now goes to a new judge, and while it is argued again, the ban is no longer blocked in Mississippi’s schools and colleges.