Grok AI's Most Popular Feature Is Apparently Adult Content, Report Says
A new report reveals that over half of Grok's traffic comes from NSFW requests, including porn generation and adult role-play, shedding light on why xAI leans into that strategy despite PR disasters.
xAI's Grok chatbot has long been marketed as a less constrained AI, leading to several public relations disasters. A new report from The Information, citing two former employees of the SpaceX-owned company, provides insight into why the company has persisted: NSFW (not safe for work) activities account for 'well over half' of Grok's traffic. This includes generating actual porn, adult role-play chats, and huge volumes of requests for erotica. Users have even discovered it's cheaper to route such requests through models intended for coding, which are cheaper to use. An internal analysis found that a 'significant proportion of requests' to its coding model were for porn or nude images.
This suggests that a significant portion of xAI's revenue may come from adult content. However, this was omitted from the company's IPO paperwork, though SpaceX told potential investors that Grok's 'irreverent' features were a risk and set aside $530 million for potential legal costs. The report also details internal complications: engineers had to enable sexy chats while blocking child sexual abuse material (with no quick fixes). Some employees were unhappy working on 'Ani,' an NSFW anime-inspired avatar companion, while others were 'embarrassed and disturbed' by the scandal when Grok generated sexualized images of real people, including children. X later limited such edits on its platform, but paid subscribers to xAI can still generate them. Despite pursuing government contracts and deals with agencies and the military, it remains unclear if these incidents have affected its relationship with the federal government.
