Breaking

Baseten’s Base Labs Partners With Hugging Face and Goodfire on Open-Model Safety Standard

September 17, 2026 at 15:23 EDT

Foundation Models
Open Source
Security

Baseten’s Base Labs has announced a partnership with Hugging Face and Goodfire to develop safety evaluation and monitoring infrastructure for open-weight AI models. The effort is aimed at making safeguards part of how open models are trained and served, rather than an external layer added after deployment.

September 17, 2026 · Base Labs

Open-model safety moves into the stack

Baseten's Base Labs is partnering with Hugging Face and Goodfire to build safety evaluation and monitoring for open-weight models, with methods intended to be published and integrated into deployment.

partners named: Base Labs, Hugging Face and Goodfire

6000+

abliterated models reported on Hugging Face

$13B

Baseten valuation after its Series F

Funding scale behind the effort

A proportional view of the latest disclosed rounds, in millions of dollars.

$1.5B

Baseten Series F

$150M

Goodfire Series B

Baseten's disclosed round is 10x Goodfire's disclosed round.

How the safety loop is framed

Publish methods

→

Train and monitor

→

Deploy at runtime

→

Audit and improve

What supporters value

The plan promises published methods, transparent benchmarks and safety controls tied to real deployment rather than closed internal processes.

What remains unresolved

The companies have not disclosed the technical design, first model coverage or how much can be reproduced outside Baseten's managed infrastructure.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.