Seeing is believing the real and present threat of fake AI images – Winnipeg Free Press

Opinion

We are in a new era of sexual harassment.

At the end of January, Taylor Swift's sexy videos created by artificial intelligence began to spread on social networks. These fake images have been viewed millions of times and may have come from the popular 4chan message board as part of an “invitation” to bypass security measures and filters, according to a new report published this week. The New York Times.

Swift's images have prompted politicians to move faster when it comes to legislating the distribution of intimate videos without consent and creating them now.

If you remember, in 2014, private nude photos of celebrities were stolen from their iCloud accounts and leaked all over the internet.

Public reaction at the time was largely unsympathetic; These celebrities “asked for it” in the first place by being famous and taking these photos. The cultural rights of celebrities have been fully demonstrated.

Now, 10 years later, computer literate people don't have to go to the trouble of hacking someone's account and stealing their personal photos to harass and abuse. They can just do it. This should be very important to anyone who uses the internet.

These fake pictures are no less damaging to Swift, as she is currently the most famous woman in the world and has access to a powerful base of Swifties who can remove them from social media. Such behavior does not exist in a vacuum. What happens to pop stars also happens to teenage girls.

In December, Winnipeg police investigated reports of AI-generated nude images of underage students circulating at Beliveau College, a grades 9-12 high school in Windsor Park.

Take a moment to imagine how it would feel to have your likeness used in this way. Imagine how weak and broken you would feel.

I've seen quite a few threads on the internet about the “credibility” of Taylor Swift videos, usually run by people for whom deep fake porn harassment is just a topic of discussion.

First, if we were talking about the crude cutout of Taylor's head, assembled with scissors and a glue stick and glued to the body of a Playboy medium without her consent, it would still be wrong.

Second, whether we're talking about a celebrity, a 10th grader, or a woman in the marketing department, it doesn't matter if the images “look real.” They are real in every sense – and will live forever knowing the Internet. The damage they cause is also very real.

In addition, AI technology is rapidly improving. Along with declining media literacy, people may also stop questioning what they see.

The speed and ease with which these images are created and distributed is also troubling; you don't need to master photoshop anymore.

However, despite the rapid pace of technology, we still seem to be stuck in 2014 when it comes to the law. According to a recent Canadian Press story about Canadian provinces playing on the file, Manitoba is actually one of eight provinces with intimate image laws, but ours does not refer to altered images.

This needs to change quickly and quickly.

We cannot enforce the creation and distribution of AI-generated sexual content in the traditional way of combating online sexual harassment, which is to tell women to “stay off the internet.”

We are still online. Shouldn't we have agency over how we look?

[email protected]

Jen Zoratti
Appraiser

Jen Zoratti is a Winnipeg Free Press columnist and author of the NEXT newsletter, a weekly look at the post-pandemic future.

Read full bio