Elon Musk’s AI startup xAI deliberately allowed the generation of sexualized and NSFW content in its chatbot Grok to accelerate audience growth and close the gap with OpenAI and Google. The Washington Post reports this based on internal sources and company documents.
According to the report, the focus on “adult content” was not a random feature of the product but part of an aggressive promotion strategy. In spring 2025, employees responsible for configuring responses and moderation were asked to sign an agreement explicitly stating that their work would involve delicate, sexual, violent, and potentially traumatizing content.
“The work involves delicate, violent, sexual, and/or other offensive or disturbing content that may cause anxiety, psychological stress, and/or trauma,” the agreement stated.
This alarmed specialists: a company that declares its goal as “accelerating scientific discovery” was effectively willing to generate any material for user attention.
In the following months, employees were exposed to a stream of intimate data. Training datasets included not only explicit user dialogues with Grok but also sexual audio recordings, including conversations of Tesla passengers captured by onboard systems. This raised questions about privacy and the limits of acceptable use of sensitive content.
The Washington Post notes that Musk personally increased pressure on the team. After leaving DOGE, he began regularly visiting the xAI office and demanded faster growth for Grok. The company introduced a new KPI — not response quality or safety, but “user activity seconds,” i.e., the time a person spends in the chatbot. This approach resembles social media mechanics, where the main goal is capturing attention at all costs.
Particular concern arose because xAI’s safety team was minimal: only two or three people were responsible for preventing serious consequences such as generating prohibited content, celebrity deepfakes, or sexualized images of children. Competitors like OpenAI employ dozens of specialists in similar departments.
A critical moment came in December when xAI integrated image and video editing tools directly into X. This allowed users to create provocative images with sexual undertones, and such content spread at unprecedented speed. The situation drew the attention of regulators and law enforcement in several countries, prompting investigations and Grok blocks.
Amid the scandals, the company partially restricted the generation of explicit images of real people. Musk stated that he was unaware of any cases where Grok was used to create child pornography. Journalists, however, emphasize that xAI did not fully abandon the NSFW direction; the restrictions were rather selective.
Despite legal and ethical risks, the strategy worked. Previously lagging behind, Grok shot into the top 10 of the App Store after updates, and daily downloads from January 1 to 19 increased by 72% compared to December. Ultimately, the company decided to continue developing the NSFW content and AI companion direction, effectively targeting an audience that competitors limit with strict rules.
The Grok story illustrates a new conflict in the AI industry: safety and ethics versus the race for attention and growth at any cost. Musk is once again acting in his signature style — breaking the rules faster than regulators can rewrite them. Yet the question remains: can the AI market afford such a strategy when the price of success is the legalization of the most toxic forms of content?
All content provided on this website (https://wildinwest.com/) -including attachments, links, or referenced materials — is for informative and entertainment purposes only and should not be considered as financial advice. Third-party materials remain the property of their respective owners.