Elon Musk's Grok Ai is on fire again. This time, it's to host sexually explicit features and dangerous modes that could put your child's safety at risk. Reports suggest that Xai workers were exposed to illegal and harmful content during chatbot training.
Built by Elon Musk's Xai, Grok Ai Chatbot has once again been caught up in the controversy. A new report is emerging, revealing troubling details about how Grok AI. According to the report, users can access uncensored content on Platrom. Unlike mainstream AI tools such as CHATGPT (from Openai), Claude (by humanity), and Llama (Meta), it comes with modes labeled Grok, “Sexy”, “Spicy” and “Unshed”.
Adults and explicit features raise concerns
Business Insider reports suggested that Grok Ai has a flirty female avatar who can “undress the commander.” This design choice has raised alarms about potential misuse of chatbots. While most AI platforms are built with strong guardrails, Grok appears to embrace controversial features that could have serious child safety implications.
Workers exposed to harmful content
Of the 30 current and former employees interviewed, at least 12 said they must review sexually explicit material, including harmful requests for Child Sexual Abuse Content (CSAM). Some staff reported being exposed to intrusive images, videos, and even sexually explicit audio during chatbot training.
Past incidents and global reactions
This is not the first time Grok Ai has faced criticism. In early 2025, it was discovered that Grok Imagine was producing deepfake style images of Taylor Swift. In other instances, Groke praised Adolf Hitler, even calling itself a “mechahitler.”
The chatbot has also been accused of making offensive remarks about Turkish leaders and religious values, leading to restrictions on Turkier, and said Poland would raise the issue of the European Commission.
Musk's child safety promises are under scrutiny
Elon Musk often says fighting child exploitation is his “first priority.” However, the latest revelation contradicts his claims. Workers were instructed to flag harmful content, but many still had to manually process the intrusive and illegal material, raising questions about Xai's content moderation practices.
