OpenAI glitch locks Microsoft employees out of ChatGPT

Almost a year since ChatGPT gained global popularity, its success has drawn widespread attention. Despite its acclaim, concerns about data security have arisen, particularly as organisations observed their employees inputting company information into the AI chatbot for task completion. Consequently, several organisations, especially those handling sensitive data, opted to either restrict or prohibit the use of ChatGPT.

Samsung was among the early adopters of such measures, banning its employees from utilising ChatGPT after discovering instances of uploading sensitive code to the platform. Any data fed into ChatGPT becomes publicly accessible, prompting heightened caution from entities dealing with confidential information, including banks and regulated industries.

Recognising the security implications, OpenAI responded by introducing an Enterprise version, designed to prioritize the privacy and security of its users. However, the move towards more secure AI practices has not prevented occasional incidents, as evidenced by a temporary block on ChatGPT usage by Microsoft employees in November. This decision, attributed to security and data concerns, was later deemed a mistake resulting from a system test for large language models.

Microsoft, a significant investor in OpenAI, has consistently increased its commitment to the partnership. Despite the glitch, both companies remain dedicated to collaborative technological advancements. Yet, the incident fuelled speculations and rumours on social media, with OpenAI CEO Sam Altman promptly dispelling claims of any retaliatory actions against Microsoft. While the episode appears to be an isolated error, its recurrence would likely lead to more significant interpretations by observers.