team-pcp-sells-mistral-ai-repositories-under-minishaihulud-pressure

Post Views: 1

A threat actor linked to the TeamPCP identity has come forward claiming to sell alleged internal repositories and source code from French artificial intelligence company Mistral AI on a popular hacking forum.

The Claimed Repositories

The claim surfaced just days after the Mini Shai-Hulud supply chain attacks on npm and PyPI packages associated with Mistral AI. The threat actor, who used the TeamPCP name, advertised roughly 5GB of alleged internal repositories connected to both “mistralai” and “mistral-solutions.”

Training systems
Fine-tuning projects
Benchmarking tools
Dashboards
Inference infrastructure
Experiments
Future AI projects

Examples of repository names listed in the post include “mistral-inference-internal,” “mistral-finetune-internal,” “chatbot-security-evaluation,” “devstral-cloud,” and “pfizer-rfp-2025.”

Risks and Concerns

While the legitimacy of the claimed repositories has not been confirmed, the list includes dozens of repository names consistent with internal engineering environments and enterprise AI development workflows.

According to the threat actor, the repositories would otherwise be leaked publicly within a week if no buyer is found.

The threat actor is requesting $25,000 in exchange for the data.

Implications and Next Steps

This development raises concerns that attacks targeting AI software environments may be expanding beyond poisoned packages and stolen credentials.

As AI companies build cloud-hosted training, inference, and autonomous agent systems, developer credentials and CI/CD environments become increasingly valuable targets for groups seeking access to intellectual property and enterprise infrastructure.

Mistral AI has yet to comment on the claims, and t