Taylor Swift seeks to shield her image from AI, but ‘deepfakes’ go faster than the law

In a digital ecosystem where the border between reality and simulation is increasingly blurred, relevant personalities such as singer Taylor Swift try to protect unauthorized use of her image with artificial intelligence (AI) generative through legal means, however, cybersecurity experts have warned that these barriers are not enough to fight deepfakes.

On April 24, through her company TAS Rights Management, the singer presented three applications to the Patent and Trademark Office of the United States to register your voice and image as a trademark, with the aim of Protect against misuse of your image and sound for generative AI.

Specifically, two of the requests were audio samples intended to protect your voice: one is “Hi, I’m Taylor Swift” and the other is “Hi, I’m Taylor.” The third, however, was visual, using a photograph Swift on stage during The Eras Tour for protect your image.

This movement responds to a series of events in which Swift’s image has been a recurring target of misuse through AI, from Meta chatbots to pornographic content that has spread through the networks. Likewise, during the 2024 presidential campaign, the president of the United StatesDonald Trump, spread AI generated images that suggested, misleadingly, that the artist supported his candidacy.

BEYOND THE ‘COPYRIGHT’

When trying to register its characteristics as a brand, the artist seeks cover the legal gap where traditional ‘copyright’ fails to protect the style or essence that AI algorithms do imitate.

That is, the AI ​​generates new content that imitates the artist without copying a previous recording. However, Swift tries to use the trademark “confusion” standard to be able to legally claim against synthetic imitations.

Therefore, by registering her iconic costume or pose, the artist confronts the abuse of her aesthetic with unauthorized avatars, also marking a roadmap for other public figures who also want to protect their image.

IT IS NOT ENOUGH AGAINST ‘DEEPFAKES’

However, experts in cybersecurity and the use of AI from the company Check Point Software Technologies, have warned that this legal shield “will not be enough to stop the direct impact of ‘deepfakes'” that Swift has been facing, given that the law “is reactive versus the speed of the algorithm”.

“Although these legal actions are historic steps,” spoofing technology requires real-time responses, like him content marking or ‘watermaking’, in order to prevent “the identity of celebrities from being used as a weapon of mass misinformation,” the company has detailed.

This is because the Registering a trademark does not prevent a ‘deepfake’ from being created as such, but, really, it only allows suing the person or company that owns the AI ​​that generated the ‘deepfake’ “once the damage is done.”

Taking this into account and given the speed with which AI acts, the ‘deepfake’ with Taylor Swift “the damage spreads faster” than any intervention by a court in this regard, as highlighted by the technical director of Check Point Software for Spain and Portugal, Eusebio Nieva.

Thus, he has made reference to the AI ​​campaign using the singer’s image during the last US elections, which reached “millions of people before it could be countered” effectively.”

REAL PROTECTION SYSTEMS

Therefore, Nieva has highlighted that the royal protection In the face of this unauthorized use of personal images, go through a “layered approach“. That is, a set of detection systems, mechanisms that prove the origin of the content y user education so that they are able to identify manipulations.

“With new elections on the horizon, the challenge is not only technological or legal, but also public confidence in distinguish what is real and what is not“said the cybersecurity expert.

With all this, Check Point is committed to, in addition to legal means, using watermarks or labels that identify that content has been generated or modified with AIavoiding sharing misleading information or creating uncertainty among users.

It is worth remembering that Taylor Swift’s case is not isolated, but rather consolidates a trend of artists and well-known figures to protect their image. It is the case of actor Matthew McConaughey, who also recently trademarked his well-known 1993 catchphrase, “All right, all right, all right,” along with other protections over your image and voice.