Optum Restricts AI Chatbot Access After Security Breach Concerns

Optum, a subsidiary of UnitedHealth Group, has restricted access to its internal AI chatbot after it was found publicly accessible online, raising security concerns. The "SOP Chatbot," intended for employee use in answering questions related to health insurance claims and standard operating procedures, was exposed without a password, though it did not contain sensitive personal data. Following a TechCrunch inquiry, Optum quickly removed the chatbot, clarifying its limited purpose as a demo tool that only referenced existing internal documents. Despite its intended functions, the chatbot had been used extensively by employees, sometimes generating responses outside its training, highlighting its unintended versatility.

Opens on TechCrunch