Inside the 48-hour Grok crisis that put X in MeitY’s crosshairs

X said that the meeting did not take place but multiple sources have confirmed to Mint the sequence of events from the meeting until the notice was issued.

Last Wednesday, the last working day of the year, X’s India leadership team was summoned to the ministry of electronics and information technology (MeitY). At Electronics Niketan, near Jawaharlal Nehru Stadium, senior bureaucrats conveyed their displeasure over how Grok—an AI platform owned by Musk and embedded into X—allowed users to manipulate images posted on the social media platform, triggering concerns over digital shaming of political leaders, a person directly aware of the development told Mint.

During the hour-long meeting, X executives explained how the image-editing feature functioned. They offered an example: if billionaire Elon Musk appeared in a photograph alongside other technology executives and users prompted the tool to remove “the most racist person", Grok might remove Musk, this person said.

The explanation failed to reassure regulators. Officials asked X to comply with local laws. Soon after the meeting ended, X’s India leadership escalated the matter to its global compliance team, an official aware of the conversation said.

X was represented by two senior executives at the meeting. Since Musk’s takeover of Twitter in October 2022, X has maintained a lean India team and does not have a designated India head following the departure of its former policy chief Samiran Gupta in September 2023.

Emails sent to X and MeitY for comment were not responded to until press time. After this story was published Sunday night, an X spokesperson wrote on email Monday afternoon that “no such meeting took place on December 31st 2025".

What followed over the next 48 hours pushed the issue beyond political satire or image manipulation, and into the realm of mass online abuse. Grok’s image-editing features were widely used to generate sexualised images of women without consent, prompting a flood of complaints and drawing regulatory scrutiny.

By Friday afternoon, MeitY had been pushed to a corner. By evening, it issued a formal notice to X. Mint has seen a copy of the notice, sent around 7 pm on Friday.

The notice, another person directly aware of the developments told Mint on condition of anonymity, took the company by surprise. “X remains in compliance with India’s legal requirements, and speaks with the ministry of electronics and IT (MeitY) every week. At no point in prior conversations were sexual content issues brought up before, as recently as two days before the notice. X’s voluntary action reports in India are filed every month, which already includes the details asked for. It remains unclear if the government wants information beyond that," the person said.

The 31 December meeting had ended without a definitive outcome, suggesting that discussions between the company and the government were still open and evolving. “AI is an evolving technology, and we’re all trying to understand what the right way of dealing with it is without stifling free speech and creative use cases," the person cited above added.

X defended to MeitY that Grok’s results are driven by reports from “credible sources and information that have not been publicly debunked, not just in India but globally." The company’s executives assured MeitY that “for any objectionable content reported beyond proactive monitoring, actions are taken on a per-case basis," the first person cited earlier said.

Misuse fallout

Yet, since 1 January 2026, misuse of Grok to generate sexualised versions of people’s photographs without consent has swept X in India. The person cited above defended the company, stating that most such content “were being generated voluntarily by users on the platform, and was not a result of Grok creating them on its own."

To be sure, Grok’s two main competing platforms—Google’s Nano Banana and OpenAI’s ChatGPT Images—do not allow users to edit or generate content where a person’s photograph is morphed into potentially objectionable imagery. Mint has independently verified this claim.

According to two people aware of the matter, this distinction led MeitY to issue a subsequent notice to X, giving the company 72 hours, until the evening of Monday, 5 January, to respond.

A senior government official involved in the discussions, who also requested anonymity, confirmed that X and xAI, Grok’s holding company and X’s subsidiary, have not yet responded. The companies may either submit their response by Monday evening or seek an extension.

“If they ask for an extension for legitimate reasons, we’ll consider that too," the official said.

“The government meets each of the social media platforms individually every week, as per clauses included in rules such as section 69 of the IT Act, 2000. These meetings are to discuss various compliance and reporting issues, and that includes a wide variety of things. The bigger question is that for users, it is important to not engage with all sorts of tools and features on a platform, which is what leads to a wider range of mischief and issues online," the official added.

For now, X’s future in India remains under scrutiny. The country is X’s fourth-largest market globally, with 22 million monthly active users as of November last year.

In its notice, MeitY has asked X to submit an “action taken report" detailing “specific technical and organizational 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; and mechanisms put in place to ensure compliance with the mandatory reporting requirement under section 33 of the BNSS (Bharatiya Nyaya Sanhita)."

Mint also reviewed X’s three most recent monthly action reports for September, October and November. In September, the company took action on 1,680 reports, of which 55% related to abuse, harassment, and sensitive adult content. In October and November, as reports related to sexual content declined, X recorded a sharp rise in complaints concerning “synthetic and manipulated media", accounting for half of 1,959 actions in October and 21% of 1,528 actions in November.

“We take action against illegal content on X, including Child Sexual Abuse Material, 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 said in a statement on its Safety handle on Sunday.

The data suggests that X was, at least to some extent, aware of rising content manipulation issues linked to Grok.

“The problems arise out of certain policies and mechanisms that Grok follows, as against some other platforms. What is permissible on X and Grok, and what isn’t, are clearly and reasonably defined in their policies—users are free to engage and report in accordance with that," the person cited above said.

Safe harbour test

However, discussions with the Centre are unlikely to align with X’s stance. The official cited earlier reiterated the government’s position: “for any kind of sexual content online, you must remove them proactively, and if you don’t, safe harbour protection is of course on the line."

Legal experts say the episode underscores unresolved ethical and regulatory questions surrounding AI-generated content.

“As far as sexually explicit content is concerned, users are well within their rights to pursue prosecution against manipulated computer imagery that may result in their pictures being converted to nude or sexual content, as India does not allow for any form of sexualized content, online. However, the time is now ripe for India to have AI specific laws and regulations specifically against deepfakes and such forms of manipulated imagery harming their reputation and dignity," said NS Nappinai, senior counsel at the Supreme Court and founder of Cyber Saathi, a cyber security advocacy body.

Nappinai said India’s proposal requiring platforms to label artificial-generated or significantly edited content may not be sufficient, and called for “specific stringent penal provisions".

She also said the definition of safe harbour protection requires greater clarity to explicitly hold platforms accountable for content generated by their own tools.

“Safe harbour exemption is meant to protect platforms against prosecution for third party content, but where such content is created or as in this case generated by the platform or tools it is offering, the platform cannot and ought not to be permitted to seek such protection. Grok is importantly a part of the platform itself, and developed and deployed by it conscious of its capabilities. It would thus be erroneous to presume that Grok’s actions should be immune to prosecution," she said.

“Merely because a platform may otherwise be an intermediary does not give it exemption for all content or actions on its platforms."

(This story was updated to add a comment from an X spokesperson.)