‘The company wouldn’t even know…’: Researcher on ChatGPT fixing security flaw that could have exposed Gmail data

OpenAI has patched a security flaw in its ChatGPT chatbot that could have allowed hackers to access users’ Gmail data. According to cybersecurity firm Radware, the vulnerability was found in ChatGPT’s Deep Research agent, a tool designed to help users analyse large amounts of information. The flaw, which was discovered in February, could have enabled hackers to siphon sensitive data from both corporate and personal Gmail accounts of users who had linked their accounts to the service. Radware, which discovered the vulnerability, said there was no evidence that attackers had exploited it. Earlier this month, OpenAI confirmed to Radware that it had fixed the flaw.