Editorial: Grok the truth monkey

Call it coincidence or the wry humour of providence, it’s a delicious irony that Grok, the AI chatbot of X, formerly Twitter, should subject the entire Indian right-wing ecosystem to an audacious fact-check the very day Prime Minister Narendra Modi used a popular American podcast to claim that he received from the Supreme Court a certificate of innocence for his role in the Godhra pogrom of 2002.

Grok was made free to all users of X last month but was tweaked in recent days to deliver unsanitised responses to questions asked by Twitter users, without exercising any of the self-censorship that marks all other AI chatbots. Indian Twitter users, especially those opposed to Modi, have been gleefully using Grok to nail the lies put out by the Indian right-wing ecosystem.

Among hundreds of such responses, Grok named Narendra Modi as a leader “often seen as India’s most communal politician, tied to the 2002 Gujarat riots,” set the record straight on the Godhra verdict (the courts only found no prosecutable evidence against Modi), acknowledged the doubts over the Prime Minister’s self-claimed educational credentials, and pointed out his reluctance to address press conferences and his penchant for fawning one-on-one interviews.

More significant than any embarrassment caused to Modi, this unexpected burst of candour from Grok has dealt a staggering blow to his supporters in social media. The bot has outed accounts that peddle fake news, named celebrities acting covertly on behalf of the right wing, and falsified its egregious attacks on personalities in the opposition, including Rahul and Sonia Gandhi.

Grok’s truth-telling commenced just the day right-wing podcaster Lex Fridman dropped his interview with PM Modi. Even by the low standards set by previous interlocutors, the podcast is cringeworthy, full of white privilege pretending to be impressed, even “moved”, by Modi’s self-absorption. The three-hour interview offers nothing new and seems more like an attempt to usher Modi into the international coalition of strongmen and private equity oligarchs taking shape under the leadership of Trump and Elon Musk.

The twist is Grok is Musk’s monkey, and one need be under any illusions that its utterances are just an exuberant exercise of AI-free speech. Previous iterations of the bot had told Twitter users that Trump, his deputy JD Vance and Musk are among those doing the most harm to America. System prompts then had to be worked into the bot’s DNA to override such inconvenient truths.

The question is why Grok has been given rein to see the truth, hear the truth and speak the truth in India now? This new creature has been unleashed at a time when its master is negotiating with the Modi government to open up India to Starlink and Tesla. Musk has shown he has no qualms about using his role as preeminent adviser to Trump to further his business interests and fascism worldwide. So, is Grok being used as leverage against the Modi government?

Even more important, what is the Indian government’s response going to be? True to type, will it lean on Twitter’s India operations to rein in the chatbot? That’s not much of a stretch for this government because it’s already poised to tame online dissent anyway. Either way, whether free speech buckles to government pressure or is used as leverage in the hands of a buccaneering billionaire, it would be a pity. It will reinforce the notion that AI is just a crowbar in the hands of its owners.