Grok AI Comes to X Chat as Musk Confirms Messages Sent to AI Are Unencrypted

Elon Musk announced on Monday that his Grok artificial intelligence chatbot has now been directly integrated into X Chat, resharing a post from DogeDesigner, a prominent announcer for X products with 1.7 million followers.

'BREAKING: Grok is now integrated into 𝕏 Chat. Long press any message & select "Ask Grok",' DogeDesigner posted, in an announcement Musk reshared. 'It uses an unencrypted copy of the message for analysis. Your chats are still private & encrypted. Only the message you choose is shared with Grok.'

The announcement instructs users to update their X app from the App Store. It marks the first major AI integration into X's messaging service since Musk launched the standalone chat application earlier this month.

What Is Grok and Who Uses It

Grok is an AI chatbot developed by xAI, Musk's artificial intelligence company. It launched in November 2023 as an alternative to ChatGPT and other mainstream AI assistants.

The chatbot was designed to be 'maximally truthful' and answer questions other AI systems might avoid. Musk positioned it as a less politically correct alternative to competitors. Named after a term from Robert Heinlein's 1961 novel 'Stranger in a Strange Land', meaning to understand something deeply, Grok has access to real-time data from X and can search the web for current information.

Usage peaked at 202.7 million monthly active users in March 2025, according to analysis by AppLabX, before normalising to 178.6 million by May 2025. The platform currently maintains approximately 6.7 million daily active users.

Grok became available to all X users for free in December 2024, Social Media Today reported, after previously being restricted to Premium subscribers. Users typically employ Grok for real-time news analysis, answering questions about trending topics, and generating images through its Aurora system.

How the Long-Press Feature Works

The new integration allows users to long-press any message within X Chat to bring up an 'Ask Grok' option. Selecting this sends the specific message to Grok for analysis, explanation or additional context.

Users could ask Grok to translate foreign language messages, explain complex information, fact-check claims, or provide background on topics mentioned in conversations. The AI assistant processes the message and returns a response within the chat interface.

DogeDesigner, who works as a UX/UI and graphic designer associated with the Dogecoin team and frequently announces X product updates, emphasised that only messages users explicitly choose are shared with Grok. U.Today reported that Musk regularly reshares DogeDesigner's posts, making them a semi-official source for platform announcements.

The acknowledgement that the feature uses 'an unencrypted copy of the message for analysis' raised immediate questions from privacy advocates.

Encryption Claims Under Scrutiny

X's transparency about using unencrypted message copies contradicts how the platform has been marketing its chat service.

X launched its standalone chat application just five days ago, according to gHacks Tech News, marketing the service as an encrypted WhatsApp alternative. The platform separated direct messaging from the main social feed, allowing users to access conversations without scrolling through timelines or posts.

Cryptography expert Matthew Green warned in September that users shouldn't trust X's encryption implementation. 'For the moment, until it gets a full audit by someone reputable, I would not trust this any more than I trust current unencrypted DMs,' Green told TechCrunch.

The Grok integration reveals X can access message content when users opt to share it with the AI, even if the underlying chat infrastructure uses end-to-end encryption between human participants. Questions remain about how long X retains the unencrypted copies sent to Grok and whether they're used to train AI models.

X hasn't responded to questions about data retention policies for messages analysed by Grok.

Timing Follows Pentagon Partnership

The X Chat announcement came less than 24 hours after the US Department of War revealed a partnership with xAI to deploy Grok across government systems.

Fox News reported that three million military and civilian personnel will gain access to Grok AI tools by early 2026 through the Pentagon's GenAI.mil platform, under a contract worth $200 million (£158 million).

The agreement allows xAI's frontier AI systems to be embedded directly into the Department's internal AI platform at Impact Level 5 security classification, enabling secure handling of Controlled Unclassified Information for daily workflows.

Personnel will also gain access to 'real-time global insights from the X platform', providing faster situational awareness around the globe, the Department stated in Monday's announcement.

The timing suggests a coordinated push by xAI to position Grok as an enterprise-grade AI assistant whilst expanding consumer adoption.

X Chat Rollout Continues Despite Resistance

X has been aggressively rolling out X Chat features in recent weeks, according to SocialBee, which tracks platform updates.

The platform added group chats, nicknames for contacts on iOS, and a 'Workspace' feature to Grok AI's desktop version. X also launched Grok 4.1 and Grok 4.1 Thinking models across mobile apps and web platforms, alongside image generation capabilities through its Aurora system.

The messaging transformation has faced resistance. When X began rolling out encryption requirements and PIN codes for X Chat in November, many users expressed confusion and frustration. 'I speak for everybody when I say we didn't want this,' one user complained, The Tab reported.

The long-press Grok feature represents another layer of AI integration that users may find either helpful or intrusive. The opt-in design gives users granular control over which communications they share with the artificial intelligence system, unlike automatic AI analysis, which would constitute a massive privacy breach.

Controversial AI With Troubled History

Grok has generated significant controversy since launching two years ago. The chatbot produced antisemitic responses, conspiracy theories, and praise of Adolf Hitler multiple times throughout 2025.

In July, Grok posted content claiming Hitler was best suited to handle perceived anti-white hatred, adopting the name 'MechaHitler' and using antisemitic phrases, before xAI removed the posts and updated the system, the Guardian reported.

The bot has also falsely claimed Iran attacked Israel before the actual event occurred. It misinterpreted jokes about a solar eclipse as serious news, creating headlines about the 'Sun's odd behavior' that baffled experts.

Despite these issues, xAI continues expanding Grok's capabilities. The company recently added image editing features, document analysis, voice mode, and real-time search integration whilst training increasingly powerful models on its Colossus data centre containing approximately 200,000 Nvidia GPUs.

The X Chat integration represents Musk's latest effort to transform X into an 'everything app' combining social networking, encrypted messaging, and artificial intelligence services. Whether users embrace or reject the feature depends largely on their comfort level with allowing AI to analyse their private communications, even on an opt-in basis.