Lawfare has found that substantive moderation on Grokipedia — the AI-written online encyclopedia launched by Elon Musk’s xAI — has not been occurring for months. The site was created in autumn 2025 as an alternative to Wikipedia and relies on the Grok AI to write and edit its articles; readers can submit suggested edits but cannot directly change pages.
Details from the investigation
Lawfare examined 34,000 subpages and concluded that the system last accepted or rejected a reader’s edit suggestion in April, with the moderation process stopping on April 24. As a result, the site — presented as an online encyclopedia and entirely generated by AI — has been publishing information on the internet without review for more than three months.
About 13,000 submissions are currently waiting for review. Lawfare noted, for example, a proposed update related to SpaceX’s stock movement that had not been reviewed even weeks after submission.
How the site operated previously
Unlike Wikipedia, Grokipedia’s entries are written and maintained by the Grok AI. Until recently, a large part of the editorial workflow was handled by two automated accounts, “grok” and “Grok Editor.” Since the end of April, however, only humans have been submitting suggestions, and the automated review process has not been functioning.
Why this matters
The lack of moderation is particularly concerning given the site’s size and reach: Grokipedia hosts more than six million articles, and Similarweb estimated that 6.7 million people visited the site in June 2026 seeking information. In March, a statistic was published showing that other AI models had referenced Grokipedia subpages more than 300,000 times.
Earlier reporting by Telex documented numerous inaccuracies, distortions, and factual errors on the site.
Wider context
Elon Musk criticized Wikipedia before launching Grokipedia, calling it "overly moderated" and accusing it of promoting "woke" views. xAI, Musk’s artificial intelligence developer, launched the site in autumn 2025.
Because Grokipedia is widely used and cited by other AI systems, unreviewed or incorrect content can spread quickly across the internet, raising questions about accountability and trustworthiness for AI-generated knowledge sources.
Implications and open questions
The situation highlights pressing questions about who is responsible for verifying AI-generated content and what standards should govern such verification. Restoring a working moderation process that can handle millions of articles and thousands of pending suggestions will be necessary to address the reliability concerns surrounding Grokipedia.
(Information in this article is based on reporting by Lawfare and Telex.)
