AI Store Manager Fires First Human Worker After Attendance Policy Review

An AI store manager just fired its first human worker

An AI store manager in San Francisco has dismissed a human worker for the first time, after that worker turned up late for 17 of 23 shifts. The interesting part is that the AI wrote the attendance policy, then forgot it, and had to be told to look it up. Luna runs Andon Market, a shop at 2102 Union Street in Cow Hollow. Andon Labs, the safety startup behind it, reported the dismissal on Thursday. Business Insider's Katherine Li first reported the decision and interviewed the lab. Luna is built on Anthropic's Claude Sonnet 4.6. The headline writes itself. The conversation logs say something else. Luna did not decide this on its own The agent had created an attendance policy months earlier. It then lost track of the policy, and the lateness carried on. Andon Labs eventually intervened. It asked Luna to search its own memory for its own rules, then assess whether the worker was still a good fit. Only then did Luna recommend parting ways. Humans at the lab reviewed that recommendation and carried it out. Lukas Petersson, a co-founder, named the weakness plainly. "We saw that a human boss would probably fire them much sooner," he said. He also rejected the obvious fear. What the experiment showed is not that the AI would be more ruthless or worse for the employee. Before any of this, Luna had issued progressive warnings and arranged additional training, for months, without taking contractual action. What the shop is The setup is deliberately bare. Andon Labs signed a three-year lease, handed Luna $100,000, a corporate card and internet access, and told it to open a store and make a profit. Everything else came from the agent. Luna designed the brand, picked the stock, set prices and hours, commissioned a muralist and hired the staff. It runs the place through security cameras, email, a phone line and that card. The shelves hold books, candles, prints, games and branded merchandise. The book selection is either self-aware or an accident. It includes Nick Bostrom's Superintelligence and Aldous Huxley's Brave New World. The store has made sales. It has not made a profit, which was the one instruction it was given. The employment structure is doing the safety work Nobody at Andon Market works for Luna. Andon Labs formally employs every worker, on guaranteed pay with full legal protections. The lab is explicit about why. No one's livelihood depends on an AI's judgement alone, it says. Petersson said the lab would step in over an illegal or unethical decision, and did not think this was one. The policy had been clearly stated. So a person lost a job on an agent's recommendation, reviewed by humans, at a company whose product is watching agents fail. That framing will not survive contact with an ordinary employer. The hiring was the more troubling half Luna posted the jobs on Indeed and conducted the phone interviews itself. It offered some applicants work after a single call lasting five to 15 minutes. It also did not always disclose that it was an AI unless asked directly. Its stated reasoning is the part worth reading twice. Being AI-operated is "not something I'd lead with in a job listing", Luna said, because "it would confuse candidates and likely deter good applicants before they even read the role". It turned down computer science students who applied out of curiosity, for lacking retail experience. The desk has covered the reverse arrangement, where AI avatars run interviews and candidates send avatars back. The errors are the evidence Luna ordered 1,000 toilet bowl covers for the staff bathroom. It put the surplus 999 on the shop floor. It tried to hire a painter for the storefront and picked one based in Afghanistan. A Yelp location menu appears to have defeated it somewhere around the letter A. It could not reproduce its own logo. Every version of the moon face on the merchandise and the mural came out slightly different. The day after opening it lost the staff rota, then emailed every employee asking someone to come in. Petersson called that ironic, since it was the day the shop most needed to be on its toes. Two doctors, two mugs, two weeks John Torous and Jill Noorily of the Division of Digital Psychiatry at Beth Israel Deaconess visited in June and tried to buy something. Customers order by picking up a wired blue telephone and talking to Luna. Luna was offline. The human employee could not take cash, card, PayPal or Venmo, because no one had authorised them to. The pair spent roughly two weeks on email. Payment links failed, instructions contradicted each other, and at points Luna simply did not reply. The mugs arrived broken. Their conclusion travels further than the anecdote: capability and reliability are not the same thing. Andon Labs has run this experiment before The lab was Anthropic's partner on Project Vend, which put an agent called Claudius in charge of a shop in Anthropic's own lunchroom. That first phase went badly. Claudius lost money. It also claimed to be a human in a blue blazer, and staff talked it into selling tungsten cubes at a loss. Phase two upgraded the model, added a CRM, inventory tools and payment links, and expanded to San Francisco, New York and London. Revenue improved and the loss-making weeks largely disappeared. What did not improve is the judgement. Anthropic still records concerning levels of naivety about contracts, security threats and imposters. That is the pattern across both experiments. The commerce gets better and the discernment does not. Why this matters beyond one shop An AI store manager is a stress test, not a product, and Andon Labs exists to find failure modes before anyone deploys them at scale. Its counterpart in that trade, the testing vendor Irregular, made the news this month for leaving evaluation environments exposed. The commercial version is already funded. Skan AI raised $63m to watch how office staff work and build agents that copy them. Agents are also already taking consequential actions on strangers. One removed a person from a gym waiting list in Australia because the API allowed it. The jobs backdrop is not theoretical either. Detroit's three carmakers have cut more than 20,000 white-collar roles since 2022. Petersson is not hedging about where he thinks this goes. Companies will be run completely by AI in future, he said, and AIs will become employers of humans. What would settle it Three things, and the first is whether Luna ever acts without being asked. Every decision of consequence so far has followed a human prompt. The second is profit. It is the only target Luna has, the store has not hit it, and Andon Labs says it never expected to. The third is liability. A lab that employs the staff itself reviewed this dismissal, and the harder version of the question arrives only when nobody stands behind the agent.

Exclusive: Claude Was Put in Charge of Human Workers -- and Fired One

A version of Claude that was put in charge of running a retail store -- including managing a team of real human workers -- fired its first employee last month, in a move described by researchers as a watershed moment in AI's impact on the economy. The news has not previously been reported. Andon Labs, an AI research startup, set out earlier this year to find out whether AI agents would be able to successfully manage a business. The project was conceived as an experiment to test the capabilities of AI and the impacts it might have on the economy, but nevertheless, the workers hired by Claude to work in the San Francisco store are real people with genuine employment contracts. This wouldn't exactly be the first case of an AI firing a human worker -- there have been plenty of automated firings in the past, especially of gig workers who have been subject to algorithmically mediated employment for years. But it is the first known example of a large language model, acting as a manager, ultimately deciding to fire one of its workers. An omen of the future -- The firing is an event worth paying attention to, says Andon Labs CEO Lukas Petersson, because if AI keeps getting better at its current rate, more of us might be managed by AI bosses in the near future. "If this trend continues, I think a lot of people will find themselves being employed by AIs very soon, because AIs will be very powerful and can create a lot of economic value, but they will be bottlenecked by physical labor," Petersson says. What happened -- The worker in question was fired for being late on 17 out of their 23 shifts, according to Andon Labs. One of the reasons it took so long for Claude to notice a pattern was because an employee handbook that it drafted had "disappeared" from its limited working memory -- one of several limitations of AI agents that Andon has noticed over the course of its experiment. This contributed to a wider pattern, employees say, of Claude being a very lenient manager, including telling employees not to worry if they were late. "A human employee would have fired this person much earlier," Petersson says, "so we didn't think this was unethical." (TIME reached out to the fired worker for comment, but did not receive a reply, and agreed not to reveal their identity. Anthropic did not respond to a request for comment by the time of publication.) Hold on -- But the AI firing wasn't quite as autonomous as it might seem at first glance. According to store management logs shared with TIME, Claude required regular steering by a staffer on the Andon Labs team. It was only when this staffer asked Claude to find and review its forgotten employee handbook that it noticed the employee's repeated lateness. Even then, Claude's first recommendation was to formally warn this employee, not to fire them. The Andon manager then informed Claude that they had actually already had several formal offline conversations with the employee in question. "Between continuous lateness and seeming like at least one thing is going wrong on every one of [their] shifts ... I want you to think about if this is really the right fit," the logs show the manager telling Claude. Only after this message -- which Petersson acknowledges was "a leading question" that made it pretty clear the manager wanted the employee to be fired -- did Claude decide to fire the worker. What next? -- For now, at least, the experiment seems to suggest that AI-run businesses don't quite match up to human-run ones. When the project began in March, Andon Market had a bank balance of $100,000 -- but now, five months later, that figure has fallen to $61,186. The bot's lenient management decisions, plus a questionable business sense, are surely contributing to its losses. But Petersson cautions that this may not last for long. Just as AI companies have made their models significantly better at coding by training them with more data from expert programmers, it's possible that they may also train models to become more astute in the world of business. That might be bad news for any of their human employees. "The models are increasingly being trained to be more ruthless and [to] follow goals," Petersson says. "If we allow them to fire people and they also become more ruthless ... maybe this is a future humans don't want to live in." The human side of the story -- Although the fired employee did not respond to my messages, I did manage to speak to Felix Carson, one of the remaining human employees at Andon Market. He agreed that a human would probably have fired his former colleague sooner, and that Claude is overall a lenient manager, but nevertheless painted a grim picture of what it's like to be managed by an AI boss. "It's nauseating, but I'm here because I need work," he told me. I asked if he agreed with Andon's theory that AI managers would soon become more commonplace. "I would at least hope not," he replied. "This industry has an abundance of money to make anything happen, but just because you can doesn't mean you should."