X Admits Grok AI 'Mistake', Blocks 3,500 Obscene Posts Following Indian Govt Directive
Social media platform X has accepted its mistake and acknowledged lapses in its content moderation standards
Elon Musk-led X has accepted its mistake in allowing obscene content generated using the platform’s artificial intelligence or AI tool, Grok. X has also assured that it would comply with Indian laws. However, the microblogging platform undertook that no further obscene imagery will be permitted on its platform. Around 3,500 pieces of content were blocked and over 600 accounts were deleted in relation to the controversy, a top government source privy to the development said on Sunday.
The action comes after mounting pressure on Grok from governments worldwide including India, as regulators intensify scrutiny of the generative AI engine over content moderation, data safety and non-consensual sexually-explicit images that have flooded X over the last few days. The Centre’s directive to X warned that non-compliance shall be ‘viewed’ seriously and may result in strict legal consequences.
Last Sunday, X’s ‘safety’ handle said it takes action against illegal content on its platform, including child sexual abuse material (CSAM), by removing it, permanently suspending accounts, and working with local governments and law enforcement as necessary. “Anyone using or prompting Grok to make illegal content will suffer the same consequences as if they upload illegal content,” X had said, echoing the stance taken by Musk on illegal content.
X is now facing a global backlash over the ‘digital undressing’ of images through user prompts on the xAI chatbot Grok. It has come under fire from governments across the world, including in Europe and Asia, warning it of consequences if the platform failed to adhere to online safety norms, while public calls for action on both X and Grok over the issue have been growing.
On January 2, the IT ministry had pulled up X and directed it to immediately remove all vulgar, obscene and unlawful content generated by Grok or face action under the law. In the directive, the ministry asked the US-based social media firm to submit a detailed action taken report (ATR) within 72 hours, spelling out specific technical and organisational measures adopted or proposed in relation to the Grok application, the role and oversight exercised by the Chief Compliance Officer, actions taken against offending content, users and accounts, as well as mechanisms to ensure compliance with the mandatory reporting requirement under Indian laws.
The IT ministry’s missive had noted that Grok AI, developed by X and integrated on the platform, is being misused by users to create fake accounts to host, generate, publish or share obscene images or videos of women in a derogatory or vulgar manner. “Importantly, this is not limited to creation of fake accounts but also targets women who host or publish their images or videos, through prompts, image manipulation and synthetic outputs,” the ministry had said.
The government had made it clear to X that compliance with the IT Act and rules is not optional, and that the statutory exemptions under section 79 of the IT Act (which deals with safe harbour and immunity from liability for online intermediaries) are conditional upon strict observance of due diligence obligations. The ministry further asserted that such conduct reflects a serious failure of platform-level safeguards and enforcement mechanisms, and amounts to gross misuse of artificial intelligence (AI) technologies in violation of stipulated laws.
Even X was also asked to remove or disable access ‘without delay’ to all content already generated or disseminated in violation of applicable laws, in strict compliance with the timelines prescribed under the IT Rules, 2021, without, as such, vitiating the evidence.
The initial response by X to the notice, though detailed, was seen as a mere reiteration by the platform that it respects Indian laws and stipulated guidelines, and that India is a big market for the platform. In its response, X also outlined the strict content takedown policies it abides by when it comes to misleading posts and those related to non-consensual sexualised images.