TrustModel.ai launches independent assessment platform for foundation models

TrustModel.ai introduced a third-party assurance platform for foundation models and AI agents used by Global 2000 companies, aiming to close gaps in safety, fairness, accuracy and compliance for open-weight models deployed in production. The service is available now and will be discussed at a Stanford summit on Oct. 1.

Why it matters: - Global 2000 companies are deploying open-weight models such as Llama, Mistral, DeepSeek, Qwen and Gemma without independent assurance for the versions running in production. - TrustModel.ai is positioning third-party AI assessment as infrastructure for enterprise AI, with evaluation, monitoring and certification designed to be verifiable. - The service is meant to help enterprises close assurance gaps in hiring, lending, healthcare and customer operations.

What happened: - TrustModel.ai announced a third-party assessment service for foundation models and the enterprises that deploy them. - The company said the platform delivers standardized evaluation, continuous monitoring and certification for AI models and agents. - The service is built to work alongside consulting firms, so enterprises can pair the platform with an advisory partner they already use. - TrustModel.ai said the service is available now for foundation-model developers, enterprises and consulting partners.

The details: - The platform operates as an AI Assurance Control Plane with three functions: evaluate, monitor and certify. - Evaluate uses automated testing across 10 trust dimensions and produces a single 0–100 TrustScore with multi-judge consensus for reproducibility. - Monitor watches deployed models and agents in production and detects drift. - Certify acts as a certificate authority for AI and issues signed, verifiable and revocable certificates that are automatically revoked when a model or agent drifts out of tolerance. - Results map to the NIST AI Risk Management Framework, ISO/IEC 42001, the EU AI Act and the OWASP Top 10 for LLM Applications, plus more than 30 sector regulations. - TrustModel.ai said it is wholly founder-funded, has no outside investors and does not build competing models. - The company said it has no reseller or channel business with model developers it evaluates, and fees are not tied to a model’s score. - TrustModel.ai said enterprises can use the platform with the consultancy of their choice. - The company is publishing its open-source model assessment methodology and a public TrustScore leaderboard on GitHub through its open-source project. - The project is intended to give enterprises, regulators and researchers a transparent, reproducible way to evaluate open-weight models against the same standards used for frontier systems. - The platform is deployed across multiple clouds and is available on Google Cloud Marketplace through the marketplace listing.

Between the lines: - TrustModel.ai is drawing a sharp distinction between frontier models that may be independently assessed by large labs and open-weight models that are fine-tuned and deployed by thousands of organizations after release. - The company is arguing that independence matters as much as technical testing, and is borrowing a financial-audit style standard to make its point. - Publishing methodology and scores publicly appears designed to make trust a shared benchmark rather than a private consulting exercise. - The move also puts consulting firms in a supporting role, with TrustModel.ai supplying the independent measurement layer.

What's next: - TrustModel.ai’s AI Assurance & Governance Summit 2026 is scheduled for Thursday, Oct. 1, at the Stanford Faculty Club. - The company said leaders from AI, enterprise and government will discuss how verifiable trust can become a competitive advantage for foundation models and autonomous agents. - Organizations can request assessments now, and consulting firms can explore partnership through TrustModel.ai.

The bottom line: - TrustModel.ai is betting that enterprise AI will need independent, continuously verifiable assurance the same way finance relies on audits.