Microsoft briefly restricted worker entry to OpenAI’s ChatGPT, citing safety considerations

Microsoft has invested billions of {dollars} in OpenAI. However for a short time on Thursday, staff of the software program firm weren’t allowed to make use of the startup’s most well-known product, ChatGPT, CNBC has discovered.

“Attributable to safety and knowledge considerations quite a lot of AI instruments are now not accessible for workers to make use of,” Microsoft stated in an replace on an inside web site. CNBC additionally seen a screenshot that confirmed that ChatGPT could not be accessed on company units.

Microsoft and OpenAI representatives did not instantly reply to requests for remark.

“Whereas it’s true that Microsoft has invested in OpenAI, and that ChatGPT has built-in safeguards to forestall improper use, the web site is nonetheless a third-party exterior service,” Microsoft stated. “Meaning you need to train warning utilizing it because of dangers of privateness and safety. This goes for every other exterior AI providers, akin to Midjourney or Replika, as effectively.”

The corporate initially stated it was banning ChatGPT and design software program Canva, however later eliminated a line within the advisory that included these merchandise. After preliminary publication of this story, Microsoft reinstated entry to ChatGPT.

Many massive firms have restricted use of ChatGPT, usually to forestall the sharing of confidential knowledge. Having been skilled on in depth web knowledge, ChatGPT composes human-like responses to folks’s chat messages. The service has over 100 million customers.

Microsoft’s replace recommends folks use the corporate’s personal Bing Chat instrument, which depends on OpenAI synthetic intelligence fashions. The 2 firms are carefully tied. Microsoft has additionally been busy this yr bringing out updates to its Home windows working system and Workplace purposes that make the most of OpenAI providers, which in flip run on Microsoft’s Azure cloud infrastructure.

Earlier this week, CEO Satya Nadella appeared onstage alongside OpenAI’s Sam Altman on the startup’s first developer convention.

Following publication of this text, Altman wrote in a submit on X, “the rumors that we’re blocking microsoft 365 in retaliation are utterly unfounded.”

In January a high-ranking Microsoft engineer wrote in a discussion board that staff might use ChatGPT however suggested in opposition to coming into confidential info, Insider reported.

WATCH: OpenAI on tempo for $1.3 billion in 2023 income, up 4543% from final yr