xAI's Grokipedia, an online encyclopedia generated by artificial intelligence that Elon Musk called a "truthful and independent alternative" to Wikipedia, has not updated a single article in more than three months. Lawfare reported on August 5 that no Grokipedia page has changed since April 24, and The Verge confirmed the same finding the next day.

The numbers tell a clear story. Researchers looked at 34,519 pages that had at least one suggested edit from users, containing a total of 225,496 recommended corrections. Not a single one of those edits has been accepted or rejected since April. The live activity feed on Grokipedia's own website shows the same thing: the "Recent changes to..." section simply says "No live edits available right now."

Grokipedia launched in October 2025 with an initial set of 885,000 articles, according to The Verge. Fox Business reported that the platform crashed on its first day. A second version released in November 2025 expanded the collection to more than six million articles. According to Wikipedia's own article about Grokipedia, by late April 2026 user requests had stopped being carried out and the system was no longer making changes to articles on its own. Whether xAI has officially shut down the project is unclear.

Musk presented Grokipedia as a competitor to Wikipedia with what Lawfare described as a "Musk-inflected approach to reputable sources." His last post on X mentioning the platform was in February. Since then, neither the product nor its main backer has said anything publicly.

Lawfare broke the story of the April shutdown in its August 5 article, titled "Grokipedia Stopped Reviewing Edits in April. It Didn't Tell Anyone." The publication had previously discussed the broader topic in a December 2025 podcast episode called "Wikipedia, Ref-Working, and the Battle Over Reality," which explored Grokipedia and the question of what makes reference sources trustworthy.

How Grokipedia was supposed to work matters here. The system combined articles written by AI (specifically, large language models — the same technology behind tools like ChatGPT) with a layer where users could suggest edits. Those suggestions would then be reviewed and either accepted or rejected. That review process is what appears to have stopped. The six million articles still exist and can be read, but nothing is changing. No corrections are being processed, no new content is being generated, and the queue of user suggestions is frozen wherever it was in late April.

This is not a temporary technical problem. There has been no error page, no maintenance notice, no announcement of any kind. The platform simply stopped processing changes and did not tell its users.

The broader context here is that AI-generated reference works are still an untested idea. Wikipedia built its reputation over two decades through transparent edit histories, sources that anyone could verify, community governance, and a huge number of volunteer editors. Grokipedia tried to compress that process into a few months, using AI to quickly produce a large number of articles and a lightweight review system to check for accuracy. Whether that approach can sustain a living, up-to-date encyclopedia was always the open question. Three months of silence does not settle it for certain, but it does suggest the commitment to keeping the project running was thinner than the launch made it appear.

What makes this worth flagging is the lack of communication. Products get paused or shut down at every technology company — that is normal. What is unusual is that Grokipedia continued to present itself as a working, living encyclopedia while its core update and review systems had effectively stopped. Users who submitted suggested edits over the past three months had no way of knowing their contributions were going nowhere.

For anyone watching the AI content space, the Grokipedia stall is a useful example of the gap between how fast something can launch and how long it can be kept running. Building a collection of six million AI-generated articles in a matter of weeks is technically impressive. Keeping it current, accurate, and actively maintained is a completely different challenge, and it is the one that appears to have gone unaddressed.