Grokipedia Edit Queue Frozen for 3 Months, Report Finds
xAI's AI-built encyclopedia Grokipedia has gone three months without processing a single suggested correction. That matters because the project sold itself as faster and more accurate than Wikipedia, yet its self-correction system now appears offline just as the site crossed six million articles.
What actually happened
Lawfare examined 34,519 Grokipedia pages carrying at least one suggested edit, spanning 225,496 recommended changes, and found none resolved within the last three months. 'In our subsequent analysis... we found no accepted or rejected corrections dated within the past three months,' Lawfare wrote, as reported by bonuz.xyz. The Grokipedia live page now displays 'No live edits available right now' where activity logs once appeared. xAI launched v0.1 in October 2025 with 885,000 entries, then expanded to more than 6,000,000 articles with v0.2 in November 2025. The Verge dated the last confirmed change to 24 April 2026. xAI had not responded to requests for comment as of publication, and Elon Musk's most recent public comment on Grokipedia came in February 2026.
How we got here
Musk introduced Grokipedia as a correction to what he called Wikipedia's ideological slant, betting that AI generation and AI review could outpace volunteer editors. The article count backs that speed claim, six million entries in about a year. Wikipedia still depends on thousands of human editors clearing edits daily. Grokipedia's pipeline was meant to replace that labor with automated review. Instead, according to Lawfare's audit of tens of thousands of flagged pages, that review appears to have stopped right as the encyclopedia hit its largest size.
Why this matters for you
For anyone using AI tools that cite Grokipedia, unresolved errors across 225,000-plus flagged edits stay live indefinitely. That is a direct risk for builders feeding reference data into AR glasses or voice assistants, since a frozen source can quietly serve outdated facts. For xAI, the gap between launch promises and current inaction weakens the pitch that AI review beats human editing at scale. Anyone weighing Grokipedia as ground truth for a wearable or agent should treat its data as unverified until the correction queue restarts.
The bigger question
Can an encyclopedia written and checked entirely by machines stay reliable without ongoing human or automated correction? Grokipedia scaled from 885,000 to over 6,000,000 articles in about a year, yet its edit backlog sits frozen. If a six-million-article AI reference cannot keep pace with its own flagged corrections, what happens when AR devices and voice assistants start citing it as fact?
What to watch
xAI has not announced when Grokipedia's edit review will resume, nor issued a public response to the Lawfare findings. Watch the site's live 'Recent changes' feed for renewed activity, and watch for any statement from xAI or Musk. bonuz.market will keep tracking how AI knowledge bases handle accuracy as they get pulled into AR and smart-glasses tools.
