Technology & AI

Mistral publishes trillion-parameter AI model's weights despite cyberattack concerns

French firm Mistral posted its Large 4 model on Hugging Face after a three-week preview, despite earlier warnings about misuse.

By Standard Topic. AI-generated from the reporting listed under Coverage. Published . Updated .

This story is no longer being updated. It reflects reporting up to .

What happened

On Thursday, October 29, Mistral released the weights of Mistral Large 4, which it calls Le Chonk, on Hugging Face. The trillion-parameter model can now be downloaded, modified and run outside the developer's control. Mistral had earlier warned that the model could help attackers carry out real cyberattacks.

Why it matters

Mistral says the open model gives organizations control over their cyber defenses that closed models cannot.

Timeline

  1. Oct 29: Mistral released the open weights of Mistral Large 4 on Hugging Face, after a preview made available three weeks earlier.
  2. Oct 30: Mistral said it is still updating the model's guardrails but has not explained how those changes will reach copies organizations have already downloaded.

Coverage

More from Technology & AI

All Technology & AI news · Technology & AI story archive