For a brief time on Thursday, Microsoft employees were not able to access OpenAI's ChatGPT, CNBC has discovered. A notice on Microsoft's internal website declared, "Security and data considerations have led us to disallow certain AI instruments from use by our staff."
Microsoft has invested billions of dollars in OpenAI, but for a brief time on Thursday employees of the software company were blocked from using ChatGPT, the startup's most famous product, CNBC has learned. Microsoft issued an update on an internal website stating that, "due to security and data concerns a number of AI tools are no longer available for employees to use." A screenshot was also seen showing that ChatGPT could not be accessed on corporate devices.Microsoft and OpenAI representatives did not comment when asked. In their statement, Microsoft explained that even though ChatGPT has safeguards, as a third-party external service, caution must be taken when using it as there are risks of privacy and security posed. This warning applies to other external AI services like Midjourney and Replika as well.The company initially announced a ban on ChatGPT and Canva design software, but removed a line in the advisory that included those products afterwards. Access to ChatGPT was then restored.A number of large companies have restricted the use of ChatGPT, typically to prevent confidential data from being shared. Trained on a large bucket of internet data, ChatGPT produces human-like responses to people's messages. The service has more than 100 million users.Microsoft's update recommends using their own Bing Chat tool, which relies on OpenAI's AI models. Microsoft and OpenAI have a tight relationship, and Microsoft has been pushing out updates to Windows and Office applications this year that make use of OpenAI services, which in turn run on Microsoft Azure cloud infrastructure.CEO Satya Nadella and OpenAI's Sam Altman appeared together onstage at the startup's first developer conference earlier this week. Altman wrote in a post on X that the rumors that they are blocking Microsoft Office 365 in retaliation for the block are "completely unfounded."In January a high-ranking Microsoft engineer posted on a forum that employees are allowed to use ChatGPT but are warned against entering confidential information, as reported by Insider.
Comments