Images of the singer manipulated by artificial intelligence (AI) tools Taylor Swift It spread widely on X last week, meaning the social network and others like it Block related searches.

On Friday, social network X was flooded with manipulated images in which artificial intelligence tools were used to make the singer appear naked. It is an application known as ‘.deepfake‘ It has the capacity to raise doubts about its authenticity, given the realism of the images or videos to which it is applied.

This great spread was achieved without Taylor Swift’s knowledge or permission.and, as reported by 404 Media, can be traced to a forum on the Telegram messaging platform where users of a community shared images modified with tools like Microsoft Designer to undress women.

Searches for ‘deepfakes’ of the American singer became a trend on X, and the platform announced on Friday that its teams said:“active deletion” of all images They identified the accounts that posted these and took action.

According to The Verge, it reached 45 million views, 24,000 reshares, and even hundreds of thousands of likes and bookmarks before one of the publications began removing them.

Additionally, however, as The Verge points out, this blocking is not completely effective because draw can be made by changing the order of some words, for example ‘Taylor IA Swift’.

According to social media analyst Matt Navarra, this measure has also been adopted by Meta for its social networks Threads and Instagram.

Pornographic ‘Deepfakes’

‘Deep fake’ It mainly affects: women. Manipulation of audiovisual content starring women accounts for 77 percent of content found on dedicated YouTube, Vimeo and Dailymotion websites and channels, rising to 99 percent, according to a 2023 study by security solutions firm Home Security Heroes. pornographic ‘deepfake’ cases.

public figuresThey are also more likely to fall victim to such manipulation, given their public visibility and the number of images at their disposal. By profession, the leading actors in most ‘deepfake’ incidents are singers (58%) and actresses (33%).

Manipulation of images with artificial intelligence tools is not new, and in fact the report reveals the evolution in 2014, with the emergence of generative adversarial networks (GANs). Create synthetic images with perfect realism.

Advances in this technology in recent years have allowed its capabilities to be expanded and even offer more affordable and easy-to-use services; this has expanded their use among users who do not need special knowledge to use them.

For example, its existence in 2017 a campaign that spreads the videos pornographic starring so-called well-known peopleOften the women turned out to be fakes: The faces had been manipulated to show the faces of actresses and singers like Gal Gadot, Taylor Swift or Scarlett Johansson without their consent.

More recently and in connection with easy access to this technology, we can talk about fake nude photos of underage girls and teenagers who met in Almendralejo (Extremadura) in September last year. children.

‘Deepfake’ blurs the boundaries between reality and fiction. Great realism in people’s appearance and voiceWith his ability to damage the image of those he impersonates, he raises questions about the veracity of what is seen.