Microsoft Temporarily Blocked ChatGPT for Employees

Key Takeaways:

– Microsoft temporarily blocked employee access to ChatGPT, an AI tool developed by OpenAI, due to security concerns.
– The restriction came on the same day OpenAI announced a DDoS attack on their system, causing global outages and stalling the rollout of GPTs.
– Microsoft stated that while they have invested in OpenAI and ChatGPT has safeguards, it is still a third-party external service, posing risks to privacy and security.
– Microsoft later restored access, attributing the blockage to a mistake during a test of control systems for large language models.
– Security concerns around ChatGPT have been raised by other companies, including Apple and Amazon.
– The incident may have tarnished the successful DevDay event where Microsoft CEO Satya Nadella joined OpenAI’s Sam Altman on stage.

Gizmodo:

Photo: Justin Sullivan (Getty Images)

Microsoft and OpenAI’s bromance hit another snag Thursday. The Windows maker temporarily blocked employee access to ChatGPT, the premiere product it invested billions in from OpenAI, due to security concerns on Thursday, first reported by CNBC.

“Due to security and data concerns a number of AI tools are no longer available for employees to use,” Microsoft said in an internal update seen by CNBC. Corporate devices were temporarily unable to access ChatGPT and other AI services such as Midjourney and Replika. The restriction came on the same day OpenAI publicly announced a DDoS attack on their system, causing global outages in the week following DevDay and stalling the rollout of GPTs.

“While it is true that Microsoft has invested in OpenAI, and that ChatGPT has built-in safeguards to prevent improper use, the website is nevertheless a third-party external service,” Microsoft said internally. “That means you must exercise caution using it due to risks of privacy and security.

A Microsoft spokesperson told CNBC that the temporary blockage was a mistake resulting from a test of control systems for large language models.

”We restored service shortly after we identified our error,” said the spokesperson. “As we have said previously, we encourage employees and customers to use services like Bing Chat Enterprise and ChatGPT Enterprise that come with greater levels of privacy and security protections.”

Microsoft and OpenAI did not immediately respond to Gizmodo’s request for comment.

“The rumors that we are blocking Microsoft 365 in retaliation are completely unfounded,” said Sam Altman in a tweet Thursday night.

Apple warned employees to steer clear of using ChatGPT in May, fearing the AI could leak sensitive company information. Amazon also asked employees not to share sensitive code with OpenAI’s chatbot in January. Security concerns are widespread around the use of ChatGPT, and Microsoft would not have been the first company to block employee access to the large language model.

Microsoft CEO Satya Nadella joined OpenAI’s Sam Altman on stage during DevDay, a successful event that may have been tarnished by a week of hacks, outages, and restrictions from its largest investor.


Source link

AI Eclipse TLDR:

Microsoft temporarily blocked employee access to ChatGPT, a popular AI tool developed by OpenAI, due to security concerns. The restriction came on the same day that OpenAI announced a DDoS attack on their system, causing global outages and delaying the rollout of their language models. Microsoft clarified internally that while ChatGPT has built-in safeguards, it is still a third-party service with potential privacy and security risks. The temporary blockage was later identified as a mistake resulting from a test of control systems for large language models. Microsoft encouraged the use of its own services, such as Bing Chat Enterprise and ChatGPT Enterprise, which offer greater levels of privacy and security protections. Apple and Amazon have also previously expressed concerns about the use of ChatGPT, citing potential leaks of sensitive information. The incident comes after Microsoft CEO Satya Nadella participated in OpenAI’s DevDay event, which experienced a tumultuous week of hacks, outages, and restrictions.