Elon Musk announced that his artificial intelligence company, xAI, is developing a new feature for its chatbot Grok that will enable it to analyze online videos for AI-generated content and trace their origins across the web.

The xAI CEO made the announcement amid growing global concerns over AI-driven impersonation, deepfakes, and defamatory media, which have raised alarm among regulators and social media users.

Grok Will Analyze AI Signatures in Videos

Responding to an X user’s post warning about the dangers of AI-generated misinformation, Musk revealed that Grok will soon gain the ability to “analyze video to detect AI signatures in bitstreams” the raw digital data that make up video files and conduct “deeper web searches to determine its source.”

According to Musk, the system will be able to detect minute inconsistencies in compression, rendering, or pixel-level patterns that reveal artificial generation. Grok will also compare metadata, source paths, and digital footprints to verify a clip’s authenticity.

New Feature Follows Launch of Grokpedia

The update comes as Musk prepares to launch “Grokpedia”, a crowdsourced AI-driven knowledge platform meant to rival Wikipedia, which he confirmed will debut in two weeks.

Since its 2023 launch, Grok has been integrated directly into X, allowing users to fact-check posts and generate real-time responses. Musk has said he intends to replace X’s recommendation algorithm with Grok to deliver more personalized and verified content.

Musk’s Broader AI Vision

The feature marks another step in Musk’s campaign to position xAI as a leading alternative to OpenAI and Google DeepMind. Earlier this year, he merged xAI with the X platform, aligning the social network with his AI research ambitions.

Musk said the ability to detect AI manipulation is “essential to protect truth and authenticity in the digital age.”