xAI employee leaked an API key to private Grok models on GitHub
What happened
An xAI staff member left an API key in a public GitHub repository for about two months. It gave access to at least 60 private and unreleased models, some apparently fine-tuned on SpaceX and Tesla data. The repository was removed after GitGuardian escalated to xAI's security team.
Philippe Caturegli of consultancy Seralys spotted that an xAI technical staff member had committed a private API key to a public GitHub repository on 2 March 2025. GitGuardian's automated scanning flagged it and alerted the employee the same day, but the key still worked at the end of April. It could query at least 60 private and unreleased large language models, including unreleased Grok versions and models that appeared to be fine-tuned on data from SpaceX and Tesla. Researchers warned that anyone with it could have extracted information from those models or tried to manipulate them.
How it happened
Supply chain and access: Stolen token or key. Stolen or leaked access tokens, API keys or passwords were used to get in.
The AI involved was from xAI, specifically Grok. The product was xAI API, a type of AI building block.
Impact
Data was exposed. Data involved: credentials. The impact was confined to one organization and its users or customers.
Severity is rated High, calculated from how far the incident reached and how sensitive the data was.
Was the AI at fault?
No. The failure was in the systems, settings or people around the AI, not in the model's behaviour.
Status and timeline
Confirmed by the affected organization, the AI vendor, a regulator or a named security research firm.
It happened or began on 2 Mar 2025. It became public on 1 May 2025, 60 days later.
The repository was taken down only after GitGuardian escalated the issue directly to xAI's security team. xAI did not comment publicly.