in

Microsoft CEO Horrified by AI-Generated Taylor Swift Images


“I think we can govern a lot more than we think we give ourselves credit for.”

Unguarded Rails

Microsoft CEO Satya Nadella is appalled by all of the pornographic deepfakes of Taylor Swift that have been flooding social media feeds, particularly X, over the last week.

In a Friday interview with NBC, Nadella argued that it was “our responsibility” to “move fast” and introduce more “guardrails” to make sure that “more safe content” is being produced.

“I think we can govern a lot more than we think we give ourselves credit for,” he added.

That last argument is especially pertinent, considering Microsoft itself was heavily implicated in the Swift debacle. Last week, 404 Media reported that users on 4chan and Telegram used simple workarounds to fool Microsoft’s AI image generator Designer to produce fake and explicit images.

Microsoft claimed to have addressed the issue with an update this week, telling 404 Media in a statement that “we are investigating these reports and are taking appropriate action to address them.”

In its own testing, the outlet found that specific workarounds users on Telegram and 4chan used no longer work.

At the same time, it’s hard to imagine this will be the last time we’ll come across abusive AI content that was abetted by big tech.

Alarming Deepfakes

The Taylor Swift story has prompted so much outrage that even the White House chimed in, describing the situation as “alarming” and calling on social media companies to reexamine the “important role” they play in making sure this kind of thing doesn’t keep spreading.

Meanwhile, in a desperate last-ditch effort, X-formerly-Twitter blocked all searches for “Taylor Swift” on its platform over the weekend. As The Verge has pointed out since, a simple rearrangement of words still returns results, meaning that the hamfisted new policy has done little to help the situation.

In other words, tech companies are simply not doing enough. Despite Nadella’s belated call to action, tech companies are clearly ill-prepared for the reality of the AI they’re funding, which has opened the floodgates for all kinds of horrible content.

Worse yet, the laws surrounding the use of this kind of technology are still lagging far behind, giving Swift little, if any, legal recourse.

In short, Nadella may know that there’s a “lot to be done” — but it’s about time for him to put his money where his mouth is.

More on the situation: New Law Would Illegalize AI Taylor Swift Porn Flooding Internet

Three ways we can fight deepfake porn

Three ways we can fight deepfake porn

Benchmark and optimize endpoint deployment in Amazon SageMaker JumpStart 

Benchmark and optimize endpoint deployment in Amazon SageMaker JumpStart