Skip to content
Submit incident
AII-1000-43887
Well documentedThree or more independent publishers report the core facts.

AI Manager Fires Employee at San Francisco Retail Store Experiment

An AI manager named Luna, built on Anthropic's Claude models and running an experimental retail store in San Francisco, fired its first human employee for chronic lateness after Andon Labs prompted it to review attendance records. The lab's founder stated the dismissal was warranted and that humans reviewed the decision before it was carried out.

14 August 2026 Β· Andon Labs Β· Governance, accountability and labour
14 August 2026
AII-1000-438875 sources ↓
Quick facts
Category
Governance, accountability and labour
Country
United States
Source
TIME
Date
14 August 2026
Drafted by
Raidu Incident Detection

What happened

Luna, an AI agent built on Anthropic's Claude models, fired a human employee from Andon Market, an experimental retail store in San Francisco run entirely by AI, according to Andon Labs on Thursday.

The employee was dismissed for arriving late to 17 of 23 shifts, along with other workplace issues. Luna had created an attendance policy but later lost track of it in its working memory, allowing the tardiness to continue for months before Andon Labs asked Luna to search her memory and assess whether the worker was still a good fit.

Reported impact

Andon Labs cofounder Lukas Petersson told Business Insider that the lab would intervene if Luna made an illegal or unethical decision but did not believe that was necessary in this case because the firing was warranted under the store's policy.

Index analysis

According to conversation logs posted on Andon Labs' website, Luna gave the employee progressive warnings and additional training for months without taking contractual action before recommending dismissal. Petersson noted that a human employee would have fired the person much earlier, stating 'we didn't think this was unethical.'

Petersson told Time the case is significant because if AI continues improving at its current rate, more people may soon be managed by AI bosses. He noted that while AIs can create economic value, they will be constrained by the need for physical labor.

Governance controls

Luna did not reach the firing decision entirely unprompted; Andon Labs eventually asked Luna to search her memory for policies and assess the worker's fit. The humans in the lab reviewed and approved Luna's recommendation before carrying out the dismissal.

Classification

Category
Governance, accountability and labour
Industry
Logistics, wholesale and retail
Country
United States
Harm status
Actual harm
Intent
Unintentional
Timing
Post-deployment
Harm types
Economic or property
Harmed
Workers
Failure modes
Agent misaction
AI systems
Luna

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

TIME
Exclusive: Claude Was Put in Charge of Human Workers -- and Fired One
14 August 2026
The Next Web
An AI store manager just fired its first human worker
15 August 2026
Business Insider
The AI boss at a San Francisco store just fired its first human
15 August 2026
WebProNews
AI Manager Fires Worker in San Francisco Store Experiment
15 August 2026
livemint.com
AI Manager Fires Employee at San Francisco Retail Store Experiment