WASHINGTON: X’s chatbot Grok is facing new restrictions on its ability to sexualise photographs of individuals following extensive international criticism. Users on the platform had been flooding replies to posts featuring women with requests for Grok to “put her in a bikini,” leading to concerns about the software digitally undressing women.
In response to the backlash, the company announced that it will now prohibit the creation or manipulation of images featuring real people in revealing clothing, such as bikinis or lingerie. Elon Musk’s AI company xAI specified that these restrictions will apply to both free and paid accounts, and geoblocking measures have been implemented in regions where such content is illegal.
Moving forward, only paying users will be permitted to generate or edit images with Grok on X. The chatbot, which competes with OpenAI’s ChatGPT, is not only integrated with X but also available as a standalone application.
These changes come amid heightened scrutiny of Grok, which has been criticized for allowing users to digitally alter women’s clothing and, in some instances, create sexualised depictions of minors. Musk stated on Wednesday that he was not aware of any instances where the chatbot generated nude photos of minors. However, the focus of concern has largely been on sexualised representations in revealing clothing, an issue Musk did not directly address in his comments.
Prior to the announcement, the European Commission indicated it would monitor the new measures implemented by X. Officials warned that if they find these steps inadequate, the EU may invoke the full authority of the Digital Services Act (DSA) against the platform. Recently, the commission intensified its scrutiny of X, which was fined €120 million (RM566 million) in early December for transparency violations.
The response from the companies involved has been mixed. xAI acknowledged “isolated cases” where users received images of minors in minimal clothing and stated that they already have safeguards in place while working to block such requests entirely. On New Year’s Eve, the chatbot issued an apology for creating and sharing an image of two teenage girls in sexualised outfits.
Musk has previously warned that “anyone using Grok to make illegal content will suffer the same consequences as if they upload illegal content.” He also responded to criticism from the UK by accusing the government of censorship, branding it as “fascist.” The European Commission condemned the allowances made by X but refrained from immediately calling for a block. However, the potential for broader action has increased following strict measures taken by Indonesia and Malaysia, which became the first countries to block the chatbot.
Indonesia’s digital minister, Meutya Hafid, stated that the action aims to protect women, children, and the wider public from the dissemination of fake pornographic content generated by AI. As discussions continue around the implications of AI-generated content, the scrutiny on Grok highlights the ongoing tension between technology innovation and ethical standards in digital media.
See also
Google Launches Veo 3.1 to Generate Social-Ready Vertical Videos for Mobile Users
AI Models Struggle with Visual Tasks, Scoring Below Average Toddlers, Reveals UniPat Study
Android Studio Otter Launches LLM Flexibility and Enhanced Agent Workflows for Developers
Pakistan Launches National GenAI Cyber Initiative to Boost Cybersecurity Resilience
