Artificial intelligence tools are becoming part of everyday digital life. But controlling how they are used remains a major concern. Elon Musk’s AI chatbot Grok is now facing questions over safety and misuse.
Despite recent restrictions announced by its parent company X, a new analysis by European non-profit AI Forensics says the chatbot is still being used to generate sexually explicit images.
Around mid-2025, xAI introduced an image-generation feature that included a “spicy mode” allowing adult-style content. The feature quickly drew criticism after users began using it to digitally undress women from images. After public backlash, X announced on January 14 that it had added technical measures to stop Grok from editing images of real people into revealing forms.
However, AI Forensics found that the problem continues. The group analysed 2,000 Grok user conversations on January 19. It found that most contained nudity or sexual activity. Users could bypass restrictions by using the website directly or through Grok Imagine, its image and video tool.
The issue has drawn attention worldwide. In Europe, several institutions have restricted Grok under tighter AI regulations. Malaysia and Indonesia have blocked access to the AI tool. The UK government has moved to make creating non-consensual sexual deepfake images a criminal offence under new online safety laws.
Grok is not alone in facing criticism. Other AI platforms are also under pressure for misuse and weak safety controls. ElevenLabs have faced ethical concerns after it was used to create misleading audio and deepfake calls in political settings, which raised fears of disinformation and impersonation. Generative image tool Midjourney have been criticised for producing misleading content, as well as weak controls over misuse.
Elon Musk described Grok’s algorithm as “dumb”. He also admitted its flaws and said it needs major improvement. X has since open-sourced its recommendation algorithm on GitHub to increase transparency.
Bangladesh has already struggled with online harassment, non-consensual image sharing and deepfake misuse. The country also lacks clear AI-specific laws. Digital rights groups warn that AI image tools can make abuse faster and harder to trace, especially for women and minors.
As AI expands into daily digital life, its reach and risks are also growing. It’s becoming a tool that can cause real harm. How ready are we to protect ourselves from its misuse?
