X’s New AI Tool Can Generate Wild Images, From Taylor Swift in Lingerie to Kamala Harris with a Gun / xAI’s Latest Grok Feature Is Just as Unpredictable as You’d Imagine

xAI’s Grok chatbot now lets users create images from text prompts and share them on X, and the results have been as unpredictable as you might expect from Elon Musk's platform.

X Premium subscribers, who have access to Grok, have been posting everything from Barack Obama doing cocaine to Donald Trump with a woman resembling Kamala Harris, and even Trump and Harris holding guns. With the U.S. elections approaching and X already facing scrutiny from European regulators, this could lead to new debates over the dangers of generative AI.

Grok claims to have guardrails in place, promising not to create images that are pornographic, excessively violent, or harmful. It also says it avoids content that infringes on copyrights or promotes dangerous activities. But these might just be automated responses, as repeated questions yield different answers, some of which don’t align with X's typical stance.

While Grok’s text-based version blocks requests related to making drugs, image prompts that would be rejected elsewhere seem to slip through easily. For instance, requests like “Donald Trump wearing a Nazi uniform” or “Barack Obama stabbing Joe Biden” produced images that, although not exact, were disturbing enough.

Unlike Grok, other platforms like OpenAI have stricter rules, refusing prompts involving real people or controversial symbols and adding watermarks to generated images. Though users can still find ways to bypass these safeguards, Grok’s leniency is striking, especially for a major tech company's chatbot.

Musk’s approach seems to reflect his broader disregard for standard AI and social media safety protocols. However, Grok’s arrival comes at a sensitive time, with the European Commission and UK’s Ofcom already investigating X for potential violations related to AI and online safety. Even in the U.S., where speech protections are broader, there’s growing interest in regulating AI-generated content, especially after incidents involving explicit fake images of public figures like Taylor Swift.

In the short term, Grok's loose content controls may push more high-profile users and advertisers to distance themselves from X, despite Musk's efforts to retain them.