Grok meets Telegram: Why is AI being shoehorned into messaging platforms?

In an unusual twist, on May 28, 2025, xAI agreed to pay Telegram $300 million in cash to be integrated into the messaging platform (per Reuters). 

It is not every day you see a feature paying its platform to be combined into its existing service, but, Elon Musk is not often one for the traditional playbook. It is perhaps an astute move on both sides, purely due to the actions of competitors. AI and social media seem determined to absorb each other; from OpenAI announcing its own potential social platform, to Meta integrating an AI feature into WhatsApp and Instagram DMs. 

There does, however, seem to be some degree of circular logic to the integrations. AI companies declare AI must be big – so they put it in hard-to-ignore spots on existing services where users already spend a lot of time, which automatically results in some degree of (hard to measure) growth. Billions in investor dollars hang in the balance, as does the future of copyright. As Gzero explains, profitability is still a hypothetical, projected as an eventual guarantee, with no actual clear path to get there. Hence xAI paying Telegram to become part of the platform, and not the other way around. 

As far as brand identity goes, Telegram’s counter-culture stance and controversial CEO does match well with Musk’s X and Grok brands. The hope undoubtedly is that existing Telegram users will respond well to the integration, and users of other messaging apps may switch to Telegram as their use of integrated AI in messenger platforms becomes important enough that different AIs become a selling point. 

Yet all of this hinges on the preferences of users, and whether they actually want AI in those platforms. 

While studies have shown that the majority of time users spend on social media is spent on entertainment (e.g., Reels on Instagram), quantity of time and quality of time are two different things.  Users have been engaging more than ever in direct messages, as their friends’ presence in social feeds has been deprioritised and washed away by more marketing and creator content. MIDiA’s Consumer Survey has repeatedly found that the most important reason consumers  engage with social media is still largely to speak to and keep up with their friends, hence, the gradual retreat from the feed, to stories, and ultimately into the DMs. Adding an AI chatbot that will write their texts for them is inherently subversive of that. It degrades trust between people – did they really mean that, or did they just get AI to write something that sounds good? 

What it does do is improve the chances of monetisation for direct messaging platforms, which struggle in a way that other social platforms do not. Subscriptions for WhatsApp are limited; it is hard to put native ad placements in DMs. Integrating AI can turn them into more content-centric platforms where ads can be shoehorned in, and when, or if, users become reliant on it, it can be paywalled and ultimately become a subscription revenue driver in its own right. In the near term, it provides a use case to bring to investors as a proof of concept: “AI is being used, and is critical to the platform (because we made it that way!) More cash is needed to develop this further, and it will one day deliver the new digital world the way Facebook did back in ‘08.”

The owners of the AI platforms will defend their integrations with statements about how these add-ons “improve the experience” and “empower users to create”. But, really, what users want is very, very simple: to talk to each other. No matter the marketing, an AI bot will really only end up getting in the way – and that is something that investors and platforms alike should be much more cautious about.