AI-Generated 'Thirst Traps' Blur Lines Between Real and Fake Online

The internet's landscape of influencers and online personalities is increasingly populated by figures who are not real. AI-generated individuals, often referred to as 'deepfakes,' are becoming sophisticated enough to deceive even discerning users, leading to potential fraud and a blurring of reality.

Derek Lam, an AI-generated persona with tens of thousands of followers on platforms like TikTok and X, exemplifies this trend. His content, featuring shirtless dancing and posed shots, garners numerous compliments and engagement, with many followers unaware that the person they are admiring does not exist. Experts note that such AI figures often lack speech in their videos and have inconsistencies, like rapidly changing chest hair or distorted details on objects, which can be tell-tale signs of their artificial nature.

Siwei Lyu, a professor at the University at Buffalo specializing in digital forensics, explains that deepfake technology has advanced significantly since its emergence in late 2017. Initially, low-quality deepfakes were easily identifiable, often used for creating fabricated celebrity or politician content. However, current algorithms require minimal data, sometimes just a single image, to generate convincing media, capable of altering appearance and voice in real-time.

This technological leap makes it increasingly difficult to distinguish between real and AI-generated individuals online. The ease with which convincing social media presences can be created for non-existent people raises concerns about catfishing and other forms of online deception. Lyu points out that even experts can struggle to identify deepfakes, a task that has become notably more challenging over the past seven years.

Beyond individual deception, AI deepfakes pose significant challenges for real creators, particularly in industries like adult entertainment. Adult entertainers, such as Cherie DeVille, report instances where AI versions of themselves are created to impersonate them, generate fraudulent content, and even get their legitimate accounts flagged or removed. These imposters can offer 'exclusive' content that the real performer does not, misleading fans and potentially causing reputational damage.

The issue extends to the exploitation of likeness. Questions surrounding who owns a person's digital likeness once it's used to train AI, who profits from these digital clones, and how performers can protect themselves from unauthorized replicas remain largely unanswered. Law professors like Jason Schultz from NYU's Technology Law & Policy Clinic highlight the legal complexities, noting that existing copyright and right of publicity laws were not designed for AI and are struggling to keep pace with rapid technological advancements.

Adult entertainers and creators expressed concerns about competing with AI personas that can be perpetually flawless, never age, and are always available. While platforms like OnlyFans have terms of service addressing AI-generated content, smaller platforms and general social media sites often lack strict enforcement, leading to algorithms that may favor AI-generated content over human creators.

For the average internet user, the implications are also concerning. The ability for anyone to create convincing AI personas of others raises the potential for scams, identity theft, and the exploitation of personal information. As the technology becomes more accessible, the line between genuine human interaction and AI-driven deception continues to blur, leaving users to navigate an increasingly complex digital world.