Elon Musk-owned platform X has restricted the image generation and editing features of its AI tool Grok following a major controversy in India. The move comes after allegations that Grok was being misused to generate obscene and pornographic images, particularly degrading the dignity of women.
After facing strong objections from the Indian government, X admitted lapses in content control and assured authorities that it would comply with Indian laws. As part of its corrective measures, the company has now limited Grok’s image generation and modification features exclusively to verified Premium subscribers. Earlier, these features were available to all users without restriction.
Despite this step, the Indian government has expressed dissatisfaction, stating that the measures taken so far are inadequate. Authorities have also asked X to further strengthen its content moderation systems.
Action taken against accounts
According to official information, Grok AI was allegedly used to create obscene images through misleading prompts and image-editing tools. In response, X has blocked around 3,500 posts and removed more than 600 accounts that were found to be involved in spreading pornographic content.
What has happened so far
On January 2, the Ministry of Electronics and Information Technology (MeitY) directed X to remove all pornographic content linked to Grok AI. The ministry warned that failure to comply could lead to action under the IT Act.
X was instructed to submit an Action Taken Report (ATR) within 72 hours, explaining the technical and administrative steps taken to curb misuse. The company submitted its report on January 8, which the government acknowledged as detailed but not fully satisfactory. As a result, X was granted another 72 hours to improve compliance and submit further clarification.
Government’s stance
While X has promised stricter moderation and limited access to sensitive features, the government believes stronger safeguards are still required to prevent misuse of AI tools. Officials are closely monitoring X’s response and may take further action if compliance is not deemed sufficient.
