Company
New
Factory Joins Open Weights and American AI Leadership
August 4, 2026 - 1 minute read -
Share
Factory has signed the Open Weights and American AI Leadership letter, alongside Microsoft, NVIDIA, and 230 other organizations
Company
New
August 4, 2026 - 1 minute read -
Share
Factory has signed the Open Weights and American AI Leadership letter, alongside Microsoft, NVIDIA, and 230 other organizations

Factory has signed the Open Weights and American AI Leadership letter, alongside Microsoft, NVIDIA, and 230 other organizations. We have always built for an open ecosystem: model independence, deployment tuned to our customer controls, and open weight security models.
At Factory, we are committed to ushering in the future of AI through building a strong, open ecosystem. US leadership in AI depends on an open ecosystem that reaches every sector.
Concentrating AI behind a few closed models creates single points of failure and narrows competition. Engineering teams should choose their own models, run them where their requirements demand, and keep what they build.
Factory is model-agnostic, offering open source providers including, but not limited to, NVIDIA and Moonshot, and closed source providers.
Factory helped define AGENTS.md with OpenAI, Google, and Sourcegraph: one open file any agent can read, instead of a patchwork of vendor-specific formats. These efforts are a small subset of our work and the beginning of our commitment to grow the open ecosystem.
Defenders need models with capabilities comparable to what attackers use.
Droid Shield 2.0 uses two fine-tuned models to strengthen deterministic secret scanning. We published the LoRA adapter weights, tokenizer, configuration, training prompt, and calibration guidance on Hugging Face so other teams can run, inspect, and build on it.
Through our Open Source Commitment, we use Automated Security Review to audit open-source projects and responsibly disclose what we find, including CVE-2026-42876 in External Secrets Operator.
We are also a member of the Open Secure AI Alliance, launched with NVIDIA to build and share open tools for securing software and AI agents.
Open models still need rigor. Our Safe Autonomy Readiness Policy requires threat modeling, red teaming, sandboxing, and human approval for high-risk actions before release.
Openness and safety pull in the same direction: more people able to test, evaluate, and strengthen the systems everyone relies on. That is the ecosystem we are building for.
To learn more, reach out to sales@factory.ai.
start building
Start building