The News
Codenotary has launched AgentMon 3, the third generation of its enterprise AI runtime security platform, introducing adaptive security policies that continuously evolve based on observed AI agent behavior rather than static allow-lists. The platform currently secures more than 5 million AI agent interactions daily across enterprise deployments, giving Codenotary a substantial real-world behavioral dataset to train its policy engine. Codenotary also announced that AgentMon is now available through AWS Marketplace, reducing procurement friction for organizations already operating on AWS.
Analyst Take
The governance gap that AgentMon 3 is built to fill
Enterprise security teams are under mounting pressure to get ahead of AI agent risk before it becomes a regulatory or operational crisis. According to ECI Research’s 2026 DevSecOps + AppSec survey, AI code governance is the #1 priority investment area for enterprise security teams heading into 2026. That finding reflects a market that has moved from debating whether AI needs dedicated security controls to debating how to implement them at scale. Codenotary is betting that runtime behavioral monitoring is the answer, and the argument is credible.
The core problem AgentMon 3 aims to address is real. AI agents are not static applications. They evolve through prompt changes, model upgrades, new tool integrations, and workflow modifications. A security policy written today may be functionally obsolete in two weeks. Traditional approaches that rely on manually maintained allow-lists cannot keep pace with that rate of change, and the operational cost of trying is prohibitive. Codenotary’s claim that AgentMon can cut policy maintenance by up to 80% through adaptive, self-refining rules speaks directly to that pain point.
Why behavioral runtime monitoring matters more than perimeter controls
For developers building or operating agentic systems, the architectural significance of AgentMon’s design deserves attention. The platform monitors actual runtime behavior independently of native agent permission systems, which is an important distinction. Many AI frameworks rely on built-in permission prompts and allow-lists that developers frequently weaken or disable to reduce friction in development and testing. AgentMon detects anomalous behavior even when those native safeguards are bypassed, misconfigured, or switched off. Detection is grounded in observed file access, network activity, credential use, and process execution, not in the agent’s self-reported intent. That makes the system meaningfully resistant to prompt obfuscation and multilingual evasion techniques that fool text-based filters.
The immutable audit trail, cryptographically recorded in Codenotary’s tamper-proof ledger, adds a compliance dimension that ITDMs will find increasingly relevant. As regulatory frameworks around AI governance mature, organizations will need verifiable, tamper-resistant records of what their AI agents actually did at runtime. That is a different requirement than what most existing security tooling was built to satisfy, and it positions AgentMon directly at the intersection of security and compliance infrastructure.
The scale claim and what it signals competitively
Codenotary’s claim of 5 million daily AI agent interactions monitored is the most strategically significant number in this announcement, and not just as a marketing figure. That volume of production behavioral data creates a compounding advantage. The more interactions the platform observes across diverse enterprise environments, the more refined its threat intelligence becomes, and the harder that baseline is for a new entrant to replicate. This is a data-flywheel dynamic familiar from adjacent security markets, and it gives Codenotary a defensible position that pure-play competitors launching from a standing start will struggle to match. The AWS Marketplace listing accelerates this by lowering the deployment barrier for a large segment of enterprise buyers who prefer consolidated procurement through existing cloud agreements.
That said, ECI Research’s 2026 DevSecOps + AppSec survey also found that 67.5% of respondents selected repository access controls as a supply chain protection already in place, which illustrates that most enterprises have layered security practices anchored in earlier-generation tooling. The challenge for Codenotary is not convincing buyers that AI security matters. It is convincing security and procurement teams to add a net-new runtime monitoring layer on top of existing investments, rather than waiting for an incumbent vendor to extend its product into this space.
Looking Ahead
The agentic AI security market is in early formation, and Codenotary is moving to establish category leadership before larger platform vendors can ship credible alternatives. AWS, Microsoft, and the major cloud security incumbents will enter this space. The question is whether Codenotary can build sufficient customer depth, behavioral data, and ecosystem integrations to make displacement costly by the time those players arrive. The AWS Marketplace listing is a smart first step, but it is a distribution play, not a moat. The moat is the behavioral dataset and the adaptive policy engine built on top of it.
Over the next 12–18 months, watch for enterprise buyers to formalize AI agent governance requirements in procurement and compliance documentation. When that happens, the ability to produce a cryptographically verifiable audit trail of agent behavior at runtime will shift from a differentiator to a baseline requirement. Codenotary is positioned well for that transition. The companies most at risk are those still relying on static permission controls and manual policy reviews to govern autonomous AI systems that are, by design, never static.
Stay Ahead of Application Development Trends
Get weekly analyst insights, research notes, event coverage, and AppDevANGLE updates delivered directly to your inbox.
Subscribe for Weekly Insights
Join technology leaders, practitioners, and GTM teams following the trends shaping modern software delivery.
Looking for deeper research access?
Explore ECI Research reports, survey insights, and market analysis through the ECI Research Portal.
