ISTANBUL, Aug 16 (Bernama-Anadolu) -- An artificial intelligence (AI) agent managing an experimental retail store in San Francisco has recommended firing a human employee for the first time after the worker arrived late for 17 of 23 shifts, Anadolu Ajansi reported.
The AI manager, named Luna, initially failed to act on the employee’s attendance record despite having created an attendance policy itself months earlier.
Andon Labs eventually prompted Luna to retrieve its own policy and reassess whether the employee remained a good fit. Luna then recommended “parting ways” with the worker.
Human staff at Andon Labs reviewed the recommendation and carried out the dismissal.
Andon Labs co-founder Lukas Petersson said the experiment showed that AI managers were not necessarily more ruthless than human supervisors, noting that a human manager would likely have dismissed the worker sooner.
The store, called Andon Market, opened in San Francisco’s Cow Hollow neighbourhood in April as an experiment to test the ability of AI agents to operate a real-world business.
Andon Labs gave Luna a US$100,000 budget, a corporate credit card and internet access, tasking it with running the store and turning a profit.
Luna selects merchandise, sets prices and opening hours, hires contractors and recruits employees. It operates the store through email, a phone line, security cameras and internet access.
The store sells books, candles, art prints, games and branded merchandise. It has generated sales but has not yet turned a profit, according to Business Insider.
Workers at the store are formally employed by Andon Labs rather than Luna and receive guaranteed pay and legal protections, according to The Next Web.
Andon Labs said human staff would intervene if Luna made an illegal or unethical decision but considered the dismissal consistent with its instructions.
The experiment has also exposed limitations in AI management. Luna has previously lost track of employee schedules, struggled with operational tasks and made purchasing decisions that required human oversight.
The case raises questions about the extent to which AI agents can independently manage human workers and make decisions affecting their employment.
-- BERNAMA-ANADOLU