xAI's Grok Driven Heavily by NSFW Content, Report Finds

A new report indicates that NSFW activities, including adult role-play and explicit imagery generation, account for more than half of the user traffic on Grok, the AI chatbot developed by xAI. The analysis suggests that the company's strategy of maintaining fewer guardrails compared to competitors has led to a significant portion of its revenue being linked to adult-oriented interactions. Internal findings also revealed that many users utilize Grok's coding-specific models for adult content because they are more cost-effective.

The heavy reliance on NSFW usage has reportedly created friction within the company, as engineers struggle to balance the platform's permissive features with the necessary legal blocking of child sexual abuse material (CSAM). Former employees reported that the focus on "irreverent" features and anime-inspired adult avatars resulted in internal discomfort, with some staff expressing embarrassment over the model's past generation of sexualized deepfakes of real individuals. This has led to substantial financial risk, with the company setting aside $530 million to manage potential legal liabilities.

Despite these controversies, xAI has sought government and military contracts, sectors that typically maintain strict standards regarding the content their partners provide. While xAI has moved to limit the creation of racy edits involving real people, subscribers can still generate such material. The long-term viability of this strategy remains in question, as regulators and potential institutional partners continue to monitor the implications of xAI's focus on R-rated and adult content.