OpenAI CEO Sam Altman agreed to a bug in ChatGPT which allowed some users to see the titles of other users’ conversations on the tool. The glitch has caused concern among some firms, with some asking employees not to feed in confidential information on ChatGPT. Altman claimed the bug was caused by an error in an open-source library, but did not provide further details.
OpenAI put out a fix for the bug and temporarily shut down ChatGPT to address the issue. Altman mentioned that OpenAI will follow up with a “technical postmortem” on the incident. The latest update of ChatGPT, launched on March 14, can analyze images and have longer conversations, and is considered to be more accurate and creative than its predecessor.
ChatGPT’s versatility has brought popularity to the tool among professionals, with more than 40% having used it to accomplish tasks at work. However, some workplaces, including Amazon, Walmart, and Microsoft, have warned their employees against entering confidential information on ChatGPT over privacy concerns. Banks such as JP Morgan, Goldman Sachs, and Citigroup have also restricted their employees’ usage of the AI tool.