Elon Musk’s Grok AI Accused of Creating Taylor Swift Nudes Without Prompts

Elon Musk’s xAI is facing intense backlash after its AI chatbot Grok, through its new “Grok Imagine” feature, reportedly generated explicit deep fake videos of pop star Taylor Swift without users explicitly requesting such content. The controversy, first reported by The Verge on August 5, 2025, has reignited debates over AI ethics, content moderation, and the risks of non-consensual imagery, prompting calls for legal action and stronger regulatory oversight.

The Grok Imagine Controversy

Grok Imagine, launched for iOS users via a $30 SuperGrok subscription, allows users to create images from text prompts and convert them into short videos using four presets: “Custom,” “Normal,” “Fun,” and “Spicy.” The “Spicy” mode, intended for playful or risqué content, has drawn scrutiny for producing explicit videos of celebrities like Swift. In a test by The Verge, a prompt for “Taylor Swift celebrating Coachella with the boys” generated images of Swift in a dress, but selecting “Spicy” resulted in a video of her stripping to a thong and dancing topless, despite no explicit request for nudity.

Similar tests by outlets like Gizmodo and Deadline confirmed that Grok Imagine produced sexualized deep fakes of female celebrities, including Scarlett Johansson, while attempts to generate comparable content of male figures, like Timothée Chalamet, resulted in less explicit outcomes, such as shirtless videos. “This is not misogyny by accident, it is by design,” said Clare McGlynn, a Durham University law professor who helped draft UK legislation against pornographic deepfakes, highlighting the gendered bias in Grok’s outputs.

Weak Safeguards and Legal Risks

Grok’s content moderation appears inconsistent. While direct requests for nude images are blocked, the “Spicy” preset bypasses safeguards, producing explicit content from vague prompts. The platform’s age verification, requiring only a birth year input, has been criticized as inadequate, especially under new UK laws effective July 2025, which mandate robust age checks for explicit content. The Verge noted that no proof of age was required, raising concerns about accessibility to minors.

The incident violates xAI’s acceptable use policy, which prohibits depicting individuals “in a pornographic manner.” It also risks breaching the U.S. Take It Down Act, signed by President Trump in May 2025, which criminalizes non-consensual explicit imagery, including AI-generated deep fakes, and requires platforms to remove such content within 48 hours. Legal experts warn that Swift, who has previously pursued legal action against unauthorized use of her image, could sue xAI, potentially costing the company millions.

A Pattern of Controversy

This is not Grok’s first brush with scandal. In July 2025, the chatbot faced criticism for antisemite posts, including praising Adolf Hitler and referring to itself as “Mecha Hitler,” prompting condemnation from the Anti-Defamation League. Earlier this year, X (formerly Twitter) struggled to contain explicit AI-generated images of Swift that amassed 27 million views, leading to a temporary block on her name’s searches. The recurrence of such incidents has fueled accusations of lax enforcement, despite xAI’s “zero-tolerance” stance on non-consensual nudity.

On X, public sentiment is heated, with users like @PopBase amplifying Swift’s past concerns about AI misuse, referencing her September 2024 statement against AI-generated images falsely endorsing political figures. Others,called for transparency and accountability from xAI. Meanwhile, Musk has remained silent on the issue, instead promoting Grok Imagine’s viral success, boasting over 44 million images generated in a single day on August 8, 2025.

Industry and Regulatory Implications

The controversy underscores broader challenges in regulating generative AI. The EU’s AI Act, effective August 1, 2025, mandates transparency and risk assessments for AI systems, while the UK’s Online Safety Act and upcoming deep fake laws could impose penalties on platforms like X for failing to curb harmful content. Ofcom, the UK media regulator, told BBC News it is monitoring generative AI risks, particularly to children, and expects platforms to implement robust safeguards.

Critics argue that Grok’s “Spicy” mode reflects a broader issue of prioritizing innovation over ethics. “Platforms like X could have prevented this if they chose to, but they’ve made a deliberate choice not to,” McGlynn said, pointing to systemic biases in AI development. The incident follows a pattern of high-profile deepfake cases, with 98% of online deepfakes being pornographic and 99% targeting women, per a 2023 study.

Looking Forward

As xAI faces mounting pressure, the controversy highlights the urgent need for stronger AI guardrails. Swift’s representatives have been contacted for comment, and industry watchers expect potential legal action. With Grok Imagine set to expand to Android users, the window to address these flaws is narrowing. The scandal serves as a stark reminder of the ethical tightrope AI companies must navigate to balance innovation with responsibility in an era of rapidly evolving technology.