January 29, 2024, 12:50 pm | Read time: 6 minutes
It’s a scandal that has not only stirred Taylor Swift’s fanbase but also reached the political arena. Numerous pornographic deepfakes of the singer are flooding X, and the company has struggled for days to control them. Now there’s an update.
The fact that deepfakes and AI-generated pornography are becoming an increasingly significant problem is already well-known. However, a recent case has drawn particular attention in the media and online communities. Last week, social media platforms like X (formerly Twitter), Reddit, Instagram, and Facebook experienced a deluge of deepfake pornography featuring superstar Taylor Swift. X was forced to act quickly and temporarily blocked searches for “Taylor Swift.” Is this incompetence, unwillingness, or being overwhelmed?
Overview
Update: X Lifts Search Ban
At least the dilemma with the blocked name search has been reversed by X. “Taylor Swift” can once again be successfully entered in the image search on X. However, the company will “remain vigilant about the spread of such content and remove it wherever we find it,” assured X manager Joe Benarroch to Bloomberg. With the ban, X had attempted to temporarily prevent the search for pornographic deepfakes.
More Deepfakes
Deepfakes are images, videos, or audio recordings generated or edited using AI. With today’s technology, deepfakes often appear convincingly real. Deepfake pornography is predominantly created and distributed against the will and knowledge of those affected, making it a form of image-based sexual violence. Women and girls are particularly affected, while the perpetrators are mostly men.
Heavily promoted software offerings and nudify apps allow pornographic deepfakes to be generated for free and with low barriers. The distribution occurs almost uncontrollably, including on social networks. This was the case with Taylor Swift last Wednesday. For example, one of the deepfakes on X was viewed about 47 million times within 17 hours before it was deleted by X. On Thursday morning, X began suspending some of the accounts sharing the deepfakes. The success was moderate, as the images were quickly re-uploaded elsewhere. Reddit, Facebook, and Instagram also began deleting the deepfakes.
X Blocks “Taylor Swift” Search
In principle, X’s guidelines prohibit uploading manipulated media at all. On Friday, X’s security team announced it would “actively remove all identified images” and “take appropriate action against the responsible accounts.” X manager Joe Benarroch also said: “This is a temporary measure carried out with great caution as we prioritize safety in this matter.” The following Saturday, initial X users reported error messages when searching for “Taylor Swift.” The same applied to name combinations with the search terms “AI” (Artificial Intelligence) and “Deepfake.”
Instead of preventing the upload of deepfake pornography from the outset, X, Reddit, and Meta are now struggling to manage the consequences. They are blocking the names of those affected in the search. But even if all compromising images were to disappear from the internet forever–which is practically impossible–the damage to Taylor Swift is already done. And this applies not only to the singer but to all individuals affected by deepfake pornography.
Read also: Search engines do little against fake celebrity nude photos
Swifties Fight Against Deepfakes
According to The Verge, X’s ban could be easily circumvented by simply changing the order of search terms. This makes X’s measure appear toothless and merely symbolic. The Swifties, Taylor Swift’s fan community, acted with a strong sign–and presumably greater impact. On all social media platforms, they are trying to drown out the deepfakes by flooding the affected search terms with their own memes and images.
The pornographic content does not disappear, but at least loses some visibility on X, Instagram, and others. However, this action is unlikely to affect specific porn websites if users have already uploaded the deepfakes there.
Fans of Taylor Swift are also calling on the adult fan community to actively search for and report the deepfakes on social media. An appeal is made for Elon Musk to finally get a grip on the problem.
Case Could Stir U.S. Politics
With Taylor Swift, an exceptionally well-known celebrity is now affected by deepfake pornography. But she is neither the only one nor are only famous women at risk. At least there is a glimmer of hope: Taylor Swift’s extensive reach could indeed pressure politics and companies to address the deepfake issue and AI-generated pornography.
Democratic Congresswoman Yvette Clarke is working on legislation to combat deepfake nudes. She says: “What happened to Taylor Swift is nothing new. For years, women have been targeted by deepfakes without their consent. And with advances in artificial intelligence, it is becoming easier and cheaper to create deepfakes.”
White House Press Secretary Karine Jean-Pierre calls for Congress to draft appropriate legislation. Additionally, social media companies like X and Meta must work harder to enforce their policies “to prevent the spread of misinformation and non-consensual, intimate images of real people.” With the U.S. presidential election in November 2024, an effective tool against disinformation through deepfakes would also be desirable.
How to Check if Your Photo Is Being Used for Deepfakes
After Scandal: How X Plans to Prevent Illegal Nude Photos
X Plans to Rebuild Moderation Team
According to the European AI Act, AI-generated content–including deepfakes–must be labeled. While this regulation is welcome, it is unlikely to deter creators of deliberately fake material and pornography from their activities. Moreover, a labeled deepfake nude image is no less humiliating for those affected than a real one.
At least something seems to be happening at X: When Elon Musk took over Twitter, he laid off a large part of the moderation team. However, according to Bloomberg, 100 full-time employees are now to be hired again in the area of content control. Their main goal, however, is to combat child pornography. This is undoubtedly an important step, but it is not enough to ensure the relative safety of everyone on X.
No Statement from Taylor Swift Yet
Taylor Swift herself has not yet officially commented on the deepfakes but is said to have been “furious.” A person close to her told the Daily Mail: “Taylor’s family and friends are furious, and her fans, of course, are too.” The singer is also reportedly considering legal action against the creators of the deepfake posts. The source also clarifies: “These fake AI images are abusive, offensive, exploitative, and were created without Taylor’s consent and/or knowledge.”
Ultimately, instead of purely individual approaches, systematic and structural mechanisms are needed to significantly restrict the creation and distribution of deepfake pornography. In addition to social media companies like X and Meta, search engines also play a significant role. However, the measures of Google and Bing seem overwhelmed in the fight against deepfake pornography.