Grok spews gibberish — xAI chatbot users report 'word salad' bug across multiple sessions

Users of xAI's Grok chatbot have reported a peculiar glitch in which the model generates nonsensical, incoherent responses instead of meaningful replies. The issue came to light after one user asked the chatbot to generate a PDF and received a string of words that made no sense: "match it without and your they and two for planets can practical and often cheese…" — the gibberish continued for several paragraphs. Another user checked the source links provided by Grok and found a series of URLs pointing to reinforcement learning research sites.

Affected users who spoke to TechCrunch said they were using Grok Lite and noticed the problem as early as Wednesday morning. TechCrunch was unable to reproduce the issue in its own testing, and it is likely affecting only a small subset of users. xAI did not respond to a request for comment.

Despite the limited scope, the bug has generated a wave of confusion and frustration. The Grok subreddit has been flooded with complaints, with many users sharing screenshots of the bizarre responses. Refreshing the session often restores normal function, but some users reported that the gibberish persisted even after multiple refreshes. The bug appears to be limited to direct queries on Grok.com — the Grok account on X.com has remained unaffected.

Responding to the complaints, the official Grok account on X confirmed the issue on Thursday morning: "That pure word salad is a rare temporary generation glitch. Official status at shows all Grok services fully operational with no incidents. Start a fresh chat or regenerate—it usually clears right away. Sorry about the gibberish."

The incident comes amid reports of significant staff turnover at xAI. Earlier this year, it was reported that the company had lost most of its founding team and at least 50 researchers and engineers. The reasons for the exodus were not disclosed, but the departures raise questions about the stability of the team behind Grok. The chatbot, which was launched as a competitor to OpenAI's ChatGPT and Google's Gemini, has been marketed as a more direct and less filtered alternative. However, this glitch highlights the challenges of maintaining reliability in production AI systems, especially when key personnel leave. Users are advised to start a new chat or regenerate responses if they encounter the bug, but the underlying cause remains unclear. xAI has not provided a timeline for a permanent fix.