This document describes the security policy for the Autonomous Trust Framework for Artificial Intelligence (ATF-AI) specification and its reference artifacts.

The following versions of the ATF-AI specification currently receive security updates:

When a security issue is identified, a patch release will be published for all actively maintained versions.

If you discover a security vulnerability in the ATF-AI specification, reference implementations, or associated tooling, please report it responsibly by following the process below.

Do not open a public GitHub Issue for security vulnerabilities.

Instead, send an email to:

admin@agronet.io
Subject: [ATF-AI SECURITY] <brief description>

Include in your report:

• A clear description of the vulnerability.
• The affected component (specification section, reference artifact, or API definition).
• Steps to reproduce or a proof-of-concept if applicable.
• Your assessment of the potential impact.

We will keep you informed throughout the process and credit your responsible disclosure in the release notes (unless you prefer to remain anonymous).

ATF-AI's security model is built on the same zero-trust principles that the framework enforces for AI agent governance:

No component of an ATF-AI-governed system is implicitly trusted. Every agent action, governance decision, and execution event must carry verifiable attestations. This principle applies to ATF-AI's own specification artifacts and reference implementations.

All official ATF-AI specification releases and reference artifacts are signed with cryptographic provenance records (in-toto format). Consumers of the specification can verify the integrity and origin of any official artifact.

ATF-AI's security properties are defined at the protocol level — not tied to any specific infrastructure. This means the security model remains consistent regardless of whether ATF-AI is implemented on a blockchain network, a cloud platform, an enterprise system, or an embedded device.

All changes to the ATF-AI specification are tracked in this public repository with full version history. Any observer can audit the complete evolution of the protocol.

The following are within the scope of ATF-AI's security policy:

• The ATF-AI core specification (specs/atf-core-v1.md and future versions).
• The ATF-AI OpenAPI definition (api/openapi.yaml).
• Reference artifacts published in this repository (specs/registry.json, schema definitions).
• The ATF-AI documentation published on GitHub Pages.

The following are not covered by this security policy:

• Third-party ATF-AI adapters (e.g., erc-8040-ecosystem). Security issues in adapters should be reported to the respective adapter maintainers.
• Implementations of ATF-AI built by external organizations. AgroNet Labs is not responsible for the security of third-party implementations.
• GitHub platform-level security issues. Report those directly to GitHub.

AgroNet Labs LLC
https://agronet.ai
Security E-mail: admin@agronet.io