Red Hat’s asago translates complex regulatory policies directly into production-ready Kubernetes and cloud controls with an immutable, end-to-end audit trail.
On 4 August 2026, Red Hat, an IBM Corp. subsidiary, announced asago (AI Safety and Governance Orchestration), an open-source project converting enterprise AI governance policies into operational controls across hybrid cloud and Kubernetes environments. Building upon prior work by Red Hat and Nvidia through the Open Secure AI Alliance, asago bridges compliance, data science, and infrastructure teams.
Released under the Apache Licence 2.0 in its formation phase, asago features a public GitHub repository and community governance model. Industry collaborators include IBM Research, Microsoft, Nvidia, Alquimia AI, and Brave Software, alongside academic partners such as MIT Lincoln Laboratory, North Carolina State University, the Alan Turing Institute, IT:U Austria, and the EvalEval coalition.
Asago automates policy execution through a four-stage lifecycle:
-
Interpret & Map: Maps policy requirements to frameworks like NIST AI RMF, OWASP Top 10 for LLMs, and the EU AI Act using the IBM AI Risk Atlas.
-
Test & Recommend: Generates use-case safety tests, recommending guardrails whilst logging their rationale.
-
Deploy Configurations: Translates controls into production artifacts, including Kubernetes manifests, Terraform templates, and Ansible scripts.
-
Auditability: Maintains an immutable audit trail connecting live controls directly to policy clauses.
By replacing manual scripts, asago reduces errors as AI shifts to long-running autonomous agents. Grounded in IBM Research expertise, it enables auditors to trace active controls back to supporting evidence, with technical artifacts accessible via its GitHub repository.
