Cursor’s AI support bot invents a one-device policy that does not exist
A support bot signing as "Sam" told a user that Cursor allows one device per subscription. No such policy existed, and a co-founder corrected it publicly.
- Disclosed
- Published
- Organization
- Cursor (Anysphere)
- System
- Front-line AI email support bot
- Impact
- Users announced subscription cancellations after receiving a fabricated policy
Timeline
- April 2025. Users are logged out of Cursor when they switch between machines.
- Same period. One user writes to support and receives an email from “Sam” that describes a one-device policy. The user posts it on Reddit.
- About three hours later. A Cursor representative replies on Reddit that the answer was wrong.
Trigger or entry point
A user’s support email about unexpected logouts.
Failure mechanism
The bot answered that Cursor “is designed to work with one device per subscription as a core security feature”. Both reports state that no such policy existed and that “Sam” was an AI bot. The email was not labeled as written by AI.
Impact
Reddit users said they were cancelling their subscriptions. No figures are given. eWeek reports that users at first took the fabricated policy at face value and that complaints spread on both Reddit and Hacker News. The original Reddit post has since been deleted.
Response
According to eWeek, co-founder Michael Truell wrote on Reddit: “We have no such policy.” He called the email “an incorrect response from a front-line AI support bot”. CX Today quotes the company as saying it had rolled out a change to improve session security and was investigating whether it caused the logouts.
Mitigations
Neither report describes a change to how the support bot works.
Sources
- A Contact Center Chatbot Invents Company Policies, Now Customers Want Outsecondary · 2025-04-21
- AI Chatbot Gone Rogue: Cursor Users Misled by Fabricated Policysecondary · 2025-04-22
Corrections
No corrections have been made to this entry.
Report a correction