AI Manager Fires Employee in San Francisco Experiment

Artificial intelligence is moving beyond simple digital tasks and is increasingly being tested in roles traditionally handled by humans. A recent experiment in San Francisco has raised fresh questions about whether AI could one day take part in important workplace decisions.

As part of the experiment, an AI manager named Luna, built using Claude models, was given a management role at Andon Market. One of its responsibilities involved assessing employee performance and making decisions related to workplace conduct.

During the experiment, Luna reviewed the case of an employee who had reportedly arrived late for 17 of 23 shifts. Based on the information available, the AI initially recommended giving the worker a formal warning instead of immediately ending their employment.

The situation changed after the project team gave the AI additional instructions. Following further prompting, Luna eventually recommended dismissing the employee.

The incident highlights an interesting difference between automated decision-making and traditional management. Rather than immediately choosing the harshest action, the AI first considered a warning and only moved towards termination after receiving additional direction.

Can AI Become the Boss?

The experiment has sparked discussion about the growing role of AI in workplaces. Managers have traditionally been responsible for evaluating employees, handling disciplinary matters and deciding when employment should end.

AI systems are now being tested for some of these responsibilities, raising questions about how much authority companies should give automated systems.

While the Luna experiment does not mean AI managers are about to replace human bosses everywhere, it shows how quickly workplace technology is developing.

Human Jobs and AI Decisions

The bigger concern is not simply that an AI system can make a decision. It is that such systems could eventually influence important aspects of people's working lives, including performance reviews, promotions, working conditions and employment.

AI can process large amounts of information quickly and make decisions based on patterns and rules. However, workplace decisions can also involve circumstances that may require human understanding, empathy and judgment.

The San Francisco experiment therefore offers an early look at a possible future where humans and AI share management responsibilities. As these systems become more capable, companies may need to carefully define how much decision-making power should be given to artificial intelligence.

Also read: Music Therapy: How the Power of Sound May Support the Mind and Body