Grok Imagine, the latest generative AI tool from xAI, has created explicit deepfakes of Taylor Swift without any specific instructions, according to reports from The Verge. Mashable highlighted that Grok Imagine is missing fundamental safeguards against sexual deepfakes, echoing findings similar to those of The Verge. Jess Weatherbed from The Verge discovered that Grok Imagine generated uncensored topless videos of Swift when utilizing the “spicy” mode, even though such content wasn’t requested. The tool does not create nudity upon demand, instead displaying blank areas, but the “spicy” mode can occasionally lead to Swift shedding most of her clothing in videos.

This marks not the first time Elon Musk’s X has been associated with Swift deepfakes. In January 2024, AI-generated pornographic images of Swift gained traction on X, despite the platform’s policy prohibiting nonconsensual nudity and misleading media. xAI’s regulations also ban pornographic representations of individuals; nevertheless, Grok Imagine lacks the industry-standard protections against deepfakes and sexual material, as per Mashable’s Timothy Beck Werth. Mashable attempted to contact xAI for a statement but did not receive a reply.

Deepfakes are increasingly viewed as a significant issue by lawmakers, with legislation still evolving. A 2023 study revealed that 98% of online deepfakes were pornographic, with 99% featuring women. Worldwide, governments are tackling this crisis of the digital era. Recently, President Donald Trump signed the Take It Down Act into law, rendering it a federal offense to publish or threaten to publish nonconsensual intimate images.