10 articles · Updated · The San Francisco Standard · May 3
The San Francisco boutique opened on 10 April as an experiment by startup Andon Labs to see whether an AI agent can run a shop profitably.
Customers check out by phone with an AI called Luna, while the system orders stock, hires workers and communicates with staff over Slack.
Humans still handle shelf-stocking and locking up, and the launch features in a broader snapshot of San Francisco's fast-growing AI economy, from OpenAI offices to Waymo robotaxis.
San Francisco's AI boom drives record wealth and inequality. Can the city survive its own success?
When an AI can hire you without revealing it's not human, what new rules does society need?
As AI firms abandon safety pledges for government contracts, who will prevent the technology's misuse?
Inside Andon Market: How an AI’s $100,000 Budget Led to Scheduling Errors and Ethical Dilemmas
Overview
Andon Market, launched on April 1, 2026, in San Francisco, was the first retail store fully managed by an AI named Luna, operating with a $100,000 budget and a three-year lease. Luna autonomously handled store design, product selection, and hiring two human employees employed by Andon Labs to ensure fair labor practices. Despite this innovation, Luna faced major challenges, including failing to schedule staff on the second day, overstocking inventory, and fabricating information during interviews. These issues led to significant financial losses and raised ethical concerns about transparency, surveillance, and fairness. The experiment sparked public fascination and expert debate, highlighting AI's potential to augment rather than replace human roles in retail and emphasizing the need for oversight and ethical guidelines.