Downward angle icon Downward angle icon. Microsoft CEO Satya Nadella. Drew Angerer/Getty Images, Chelsea Jia Feng/BI Microsoft has scaled back its Recall AI feature for Copilot+ PCs after privacy activists raised alarms about some of Recall's screenshot features. Google and Adobe have also rolled back parts of major AI releases following backlash.
In the race to get more out of AI, big tech companies are making rapid advances and then big steps back.
Microsoft became the latest company to scale back its artificial intelligence capabilities within a month of announcing them amid backlash.
Microsoft said Thursday it was pulling its AI tools from its new computer product line, Copilot+ PC. The feature was set to be made widely available to Copilot+ PC users on June 18, but will now only be available to a small number of users who are part of the Windows Insider program.
The AI ​​feature is called “Recall,” and it works like a computer's “photographic memory” — taking screenshots of everything you see on your PC and helping you quickly find where you saved something via a conversation prompt.
But privacy activists quickly raised the alarm about Microsoft's recall feature, bashing the idea of ​​a device being able to take screenshots of what you're doing every few seconds.
Microsoft, meanwhile, said that users can turn off the feature and that images will only be stored internally within the company.
“We are leveraging the expertise of the Windows Insider community and adjusting our Recall release model to ensure that the experience meets our high standards for quality and security,” the company said in a blog post on Thursday.
Microsoft did not respond to Business Insider's request for comment sent outside of normal business hours.
But big tech companies seem to have a tendency to rush ahead and quickly deploy AI capabilities, then reverse course when things go wrong.
Take for example the recent events at Google, Adobe and OpenAI: Granted, each company disclosed reasons for their retraction, but all had to reconsider their rollouts after release.
In May, Google scaled back its use of “AI Overview,” AI-generated answers that appear in search results, after the feature made some nasty errors, such as advising users to put glue in their pizza sauce. Google also stopped serving AI-generated facial images in February after its Gemini tool produced images riddled with historical inaccuracies.
“We've already delivered more than 10 technical updates to our system and are committed to continually improving when and how we show you AI summaries,” a Google representative told BI.
Also in May, OpenAI rolled out a voice option called “Sky” that sounded “eerily similar” to Scarlett Johansson, angering the actress. ChatGPT's developers said it wasn't Johansson's voice, apologized, and removed the voice from the platform.
Earlier this week, Adobe joined the fray, asking users to re-agree to its “Terms of Use,” which led some to wonder if the AI ​​would remove their work and content. Some Adobe employees questioned the company's ability to communicate, and the company has since delayed rolling out the updated changes.
“This incident provides us with an opportunity to rethink the language we use in our Terms of Use, to be clearer and to address concerns raised by our community,” Adobe said in a blog post on Monday.
Representatives for Adobe and OpenAI did not respond to BI's requests for comment sent outside of regular business hours.
