Microsoft workers were prohibited from using OpenAI's ChatGPT on Thursday for a brief period of time, CNBC has learned.
Microsoft announced on an internal website that "a number of AI tools are no longer available for employees to use due to security and data concerns."
Microsoft has made billion-dollar investments in OpenAI. However, CNBC has learned that the software company's staff was prohibited from using ChatGPT, the startup's most well-known product, for a brief period of time on Thursday.
Microsoft announced on an internal website that "a number of AI tools are no longer available for employees to use due to security and data concerns." Additionally, CNBC saw a screenshot demonstrating that corporate devices were unable to access ChatGPT.
Requests for comment from Microsoft and OpenAI representatives were not immediately answered.
"The website is still a third-party external service, even though Microsoft has invested in OpenAI and ChatGPT has built-in safeguards to prevent improper use," the company stated. This implies that there are security and privacy risks, so you should use caution when using it. This also applies to any other outside AI services, like Replika or Midjourney.
After announcing that ChatGPT and the design program Canva would be banned, the company later took down a section of the advisory that mentioned those products. Microsoft reopened ChatGPT after this story's original publication.
Microsoft said in a statement to CNBC that the temporary shutdown of ChatGPT was an error that happened during a system test for large language models.
A spokesman said, "We were testing endpoint control systems for LLMs and accidentally turned them on for all employees." Shortly after realizing our mistake, we were able to restore service. We encourage staff members and clients to use services like Bing Chat Enterprise and ChatGPT Enterprise, which offer higher levels of security and privacy protection, as we have stated in the past.
A lot of big businesses have blocked access to ChatGPT, usually in an effort to stop private information from being shared. ChatGPT generates responses to users' chat messages that seem human because it has been trained on a vast amount of internet data. In excess of 100 million people use the service.
The company's own Bing Chat tool, which uses OpenAI artificial intelligence models, is what Microsoft's update suggests users use. The two businesses are interconnected. This year, Microsoft has also been quite active in releasing updates for its Office suite and Windows operating system that utilize OpenAI services, which are hosted on Microsoft's Azure cloud infrastructure.
Sam Altman of OpenAI and CEO Satya Nadella made an appearance on stage together earlier this week during the startup's inaugural developer conference.
"The rumors that we are blocking Microsoft 365 in retaliation are completely unfounded," Altman stated in a late-Thursday post on X.
According to Insider, a senior Microsoft engineer suggested against entering sensitive information in a forum post from January, even though staff members could use ChatGPT.
A hacking collective known as Anonymous Sudan claimed earlier this week that it had attacked ChatGPT due to “OpenAI's cooperation with the occupation state of Israel” and Altman's declaration that he is “willing to invest into Israel more.”