We believe AI agents need constitutional boundaries — not just permission checkboxes. We're building the governance infrastructure that makes AI-powered WordPress safe, auditable, and trustworthy.
The AOS Foundation exists to ensure that as AI agents become more powerful, the humans who deploy them retain meaningful control. Not through hope. Not through AI promising to behave. Through deterministic, verifiable policy enforcement.
AOS Governance for WP is our first CMS plugin — bringing the same constitutional governance framework we've built across the AOS platform directly into WordPress. It places a constitutional policy gate between AI agent intent and site execution. Every action is evaluated, every decision is logged, and every policy is enforced without exception.
WordPress powers over 40% of the web — making it the largest and most urgent target for AI agent governance. With WordPress 6.9 introducing the MCP Adapter and Abilities API, AI agents can now create posts, delete users, modify settings, and execute arbitrary actions on WordPress sites.
The existing protection — a permission_callback that checks WordPress roles — is
necessary but
insufficient. It answers "can this user do this?" but never "should this agent do this right
now, in this context,
at this scale?"
That gap is what we close. WordPress is the starting point — Joomla, Drupal, and other CMS platforms are on our roadmap as AI agent integration expands across the ecosystem.
Think of it like a security guard with a rulebook.
When you connect an AI agent to your WordPress site, that agent can do powerful things — create posts, delete content, change settings, even add new users. Constitutional policy governance means you write the rules before the AI acts, and those rules are enforced automatically, every single time.
It's called "constitutional" because your policies act like a constitution: they define what's allowed and what isn't, and no one — not even the AI — can override them.
Here's the key difference: Most AI safety tools either monitor what happened (after the fact) or ask the AI to follow instructions (which it can ignore). Constitutional governance blocks the action before it happens.
For example, you might write a policy that says "no AI agent can delete more than 10 posts at once" or "no AI agent can create administrator accounts." These policies are checked in under 1 millisecond, use zero AI in the decision, and produce the same result every time.
Simple rules. Absolute enforcement. Complete audit trail.
No LLM in the governance loop. No AI deciding whether to follow policy. Rules are evaluated deterministically in sub-millisecond time. The gate enforces — it doesn't think.
Every governance decision is logged with full context. Our audit trail records what happened, why it was allowed or denied, which policy matched, and the full agent context. Nothing is hidden.
We don't bolt security onto existing systems as an afterthought. We design governance as a core architectural layer — one that integrates cleanly without modifying WordPress core or your existing plugins.
The Community edition is free forever. Our governance technology is protected by patent filings, ensuring long-term sustainability while keeping the core accessible to every WordPress site owner.
Built specifically for WordPress as a first-class plugin. Uses standard WordPress hooks, APIs, and admin patterns. Zero modifications to WordPress core — installs like any other plugin and works immediately.
We believe AI governance should be accessible to everyone, not just enterprise teams. That's why our Community edition ships with real governance features — not a crippled trial. The free tier genuinely protects your site.
AOS Governance for WP is a product of the AOS Foundation, the governance technology division of Salvatore Systems. The AOS platform already provides constitutional AI governance tools including the AOS Governance Framework and production-deployed skills on the OpenClaw agentic platform. Governance for WP is our first CMS-specific deployment, with more platforms to follow.
AOS is the Agent Operating System — a framework for governing AI agent behavior through deterministic policy enforcement, immutable audit trails, and constitutional boundaries. Think of it as the governance kernel that sits between AI intent and real-world execution.
AOS Governance for WP is one piece of a larger vision for constitutional AI governance across all platforms.
Install in 60 seconds. Start governing immediately. No configuration required for the default policy set.