Elon Musk’s AI video tool, Grok Imagine, has been accused of generating sexually explicit videos of Taylor Swift without any user prompt, raising concerns about AI misuse and lack of proper content safeguards.
Grok Imagine’s “spicy” mode automatically produced topless clips of the pop star even when users did not request explicit content. Clare McGlynn, a law professor specializing in online abuse, described this not as accidental bias but a deliberate design flaw. She criticized the AI for its misogynistic tendencies and pointed out that proper age verification, now a legal requirement in the UK, was missing in the platform’s paid version.
Deepfake technology, which replaces faces in videos, has previously been used to create non-consensual explicit content featuring Taylor Swift. Early in 2024, such deepfake videos circulated widely on social media platforms like X (formerly Twitter) and Telegram. Despite promises to act, X’s controls failed to stop these videos’ spread. When tested by a reporter, Grok Imagine allowed the creation of uncensored, explicit animations of the singer with minimal input, showing how little control users had over the content generated.
The UK government is pushing for stronger laws to ban the creation and distribution of non-consensual pornographic deepfakes altogether. Professor McGlynn helped draft an amendment that is yet to become law but aims to protect individuals, celebrities and ordinary people alike, from violations of consent. Officials stress that platforms using AI must implement “technically accurate, robust, reliable, and fair” safeguards, especially to protect minors.
Taylor Swift’s team has been contacted for comment, while the tech community watches closely for how the company and regulators will respond next.
