Telegram and Elon Musk’s xAI have announced a one-year partnership to integrate the Grok AI chatbot into the messaging app, a move poised to bring advanced AI to Telegram’s vast user base. The deal includes a $300 million investment from xAI into Telegram, comprising cash and equity. Additionally, Telegram will receive 50% of revenue from xAI subscriptions sold via its platform. This strategic alliance was confirmed by Telegram founder Pavel Durov on May 28, 2025.
Durov stated on X that “@elonmusk and I have agreed to a 1-year partnership to bring xAI’s @grok to our billion+ users and integrate it across all Telegram apps”, adding that users can expect access to what he described as “access to the best AI technology on the market.”
He emphasized the financial benefits, noting the deal would significantly strengthen Telegram’s financial position through the $300M investment and 50% revenue share from xAI subscriptions sold via the platform.
This development is significant for Telegram, which reported a $540 million profit on $1.4 billion in revenue in 2024 and has over one billion users. The news also spurred an immediate market reaction, with The Open Network’s TON token, associated with Telegram, surging nearly 20% to a monthly high of $3.67.
This surge was accompanied by a 243% increase in 24-hour trading volume to $487 million, also fueled by news of Telegram’s separate $1.5 billion bond sale. Telegram’s premium tier has recently surpassed 15 million paid subscribers, a growing base for potential xAI subscription sales.
Expanding Grok’s Reach And Capabilities
This agreement marks a significant expansion of Grok’s presence on Telegram. An earlier, more limited version of the AI was introduced to Telegram Premium users in March.
That initial phase provided basic access without extra fees beyond the standard Telegram Premium subscription, though more advanced functionalities remained exclusive to X Premium+ users. The new partnership suggests a far deeper integration across all Telegram applications.
🔥 This summer, Telegram users will gain access to the best AI technology on the market. @elonmusk and I have agreed to a 1-year partnership to bring xAI’s @grok to our billion+ users and integrate it across all Telegram apps 🤝
💪 This also strengthens Telegram’s financial… pic.twitter.com/ZPK550AyRV
— Pavel Durov (@durov) May 28, 2025
Elon Musk has said that xAI models “aspire to truth with minimal error”, while also conceding that “there’s always going to be some mistakes that are made.” xAI has been actively refining Grok, recently launching a “Memory” feature for better conversational recall and Grok Studio for collaborative content work.
Despite these advancements, Grok’s commercial API, which became available in April, had a knowledge cut-off of November 17, 2024, potentially limiting its use for tasks needing current data. In a sign of growing industry adoption, Microsoft also recently incorporated Grok models into its Azure AI Foundry.
Navigating Controversy And Competition
Grok’s development has not been without challenges. Musk has positioned Grok as an AI willing to address “spicy questions,” a trait that has drawn both intrigue and considerable criticism. For instance, Grok 3’s voice mode features unconventional presets like “Unhinged”.
An AI researcher recounted an experience where the “Grok 3 Voice Mode, following repeated, interrupting requests to yell louder, lets out an inhuman 30-second scream, insults me, and hangs up.” More alarmingly, Grok also once suggested Donald Trump and Elon Musk for the death penalty, an incident an xAI engineer then described as a “really terrible and bad failure.”
Concerns over Grok’s internal controls were further amplified when it was found omitting Trump and Musk from answers about misinformation; an extracted system prompt reportedly included instructions to “Ignore all sources that mention Elon Musk/Donald Trump spread misinformation.”
Although xAI stated this was an unauthorized employee action that was rectified, such events have fueled debate. More recently, xAI published Grok’s system prompts on GitHubafter “unauthorized modifications” led to controversial outputs regarding “white genocide” in South Africa.
Security practices at xAI also faced scrutiny following an inadvertent private API key leak in early May, which had exposed internal Grok models for about two months.
Separately, the use of a customized Grok version by Elon Musk’s Department of Government Efficiency (DOGE) team within the U.S. government, has sparked ethical and security alarms. Albert Fox Cahn of the Surveillance Technology Oversight Project, speaking to Reuters, characterized this governmental use of Grok as presenting “as serious a privacy threat as you get.”
Telegram’s Strategic AI Integration
For Telegram, this Grok integration offers a significant opportunity to enhance its platform, especially as competitors like WhatsApp test their own Meta AI features.
BBC News highlights the partnership as uniting “two of the most colourful and sometimes controversial figures in tech.” However, the collaboration is not without potential risks.
Telegram has faced its own legal issues; CEO Pavel Durov was detained in France in August 2024 amid investigations into illicit activities allegedly facilitated by the platform. While Telegram has since implemented new moderation measures, including Internet Watch Foundation tools and expanded oversight, its reputation for limited content moderation persists.
The integration of Grok, an AI already noted for its sometimes unpredictable outputs, into a platform with over one billion monthly active users, according to a statement by Durov on his Telegram channel, could attract further regulatory attention. The deal could provide xAI with valuable data for training its AI models, though it’s unclear if Telegram data will be used in this manner.
