What’s Happening
Codenotary, a software supply chain security vendor, has announced two new platforms alongside a wave of enterprise customer growth. AgentMon delivers real-time observability into AI agent behavior, tracking decision chains, detecting data leakage, and monitoring token consumption across multi-agent environments. AgentX tackles a separate but related problem: automated, reversible security remediation across Linux-based infrastructure at fleet scale. The company added 37 net-new enterprise customers in six months, including Kroger, Rakuten, and Swiss Life, with its heaviest traction in defense and government. Both new products signal a deliberate strategic pivot from software supply chain integrity into AI infrastructure security.
The Bigger Picture
This announcement sits at the intersection of two converging pressures that are reshaping enterprise security priorities: the accelerating deployment of autonomous AI agents, and the persistent failure of organizations to adequately govern the infrastructure those agents run on. Codenotary is making a calculated bet that these two problems are inseparable, and the evidence suggests they’re right.
The AI Agent Security Gap Is Real and Growing
The timing of these launches is not accidental. According to ECI Research’s 2025 AI Builder Summit survey, two-thirds of enterprise AI leaders have already implemented multi-agent collaboration in live or pilot workflows. That’s a significant installed base of agentic systems operating in production today, the majority of which lack purpose-built monitoring or behavioral controls. AgentMon is a direct response to that gap.
What makes the monitoring challenge particularly sharp for agentic systems is their opacity. Traditional application monitoring tracks requests, responses, and infrastructure metrics. Agent monitoring requires something different: following decision chains, detecting anomalous delegation patterns, and catching sensitive data exposure that occurs not in a single transaction but across a distributed, asynchronous coordination loop. Codenotary’s framing of AgentMon around these specific capabilities suggests they’ve done the work to understand where conventional APM tools fall short.
The confidence question matters here too. ECI Research’s 2025 AI Builder Summit data found that 44% of enterprise AI leaders have only moderate confidence that AI agents can act autonomously without human intervention. That lack of confidence is not primarily a model quality problem. It’s a visibility and control problem. Organizations that can’t see what agents are doing can’t trust them to act alone. AgentMon, in this context, is as much a trust-building tool as it is a security tool.
What This Means for ITDMs
For IT decision-makers, the AgentX story deserves as much attention as AgentMon. Automated remediation of Linux-based infrastructure at scale aims to address an operational reality that security teams have wrestled with for years: vulnerability management backlogs that grow faster than human teams can address them.
The company’s statistic that it now secures an average of 240 compute instances per customer provides useful context. That’s a real production footprint, not a lab deployment, and it implies that AgentX is being evaluated against the performance and reliability bar that enterprise infrastructure teams apply to anything touching production systems. The “reversible fixes” framing is important. It could address the single biggest objection security automation faces in production: the fear that automated changes will cause outages. Making remediation reversible by design is a product decision that reflects an understanding of how change management works in large organizations.
The customer additions in defense and government are strategically significant. These sectors impose the strictest auditability and chain-of-custody requirements in the market. Winning there validates Codenotary’s immutable ledger approach in the most demanding compliance environments available, which makes the technology considerably more credible for regulated commercial sectors like financial services and healthcare.
What This Means for Developers and Security Engineers
For practitioners, the architecture of both products reflects a shift in where security responsibility sits. AgentX handles remediation autonomously, which may reduce the manual burden on security engineers but could also change what those engineers need to know. The skills required to oversee automated remediation are different from the skills required to perform it manually. Organizations will need to develop governance workflows around reviewing, approving, and occasionally overriding automated fixes, and that workflow design is not trivial.
AgentMon raises a related question for developers building agentic systems: what does a well-monitored agent look like at design time? The capabilities Codenotary describes, including token consumption tracking, interaction tracing, and data leakage detection, work best when the agents they monitor are designed with observability in mind. Teams deploying agentic frameworks should treat instrumentation as a first-class concern, not a retrofit.
ECI Research data reinforces why this matters now. According to our analysis, 59% of organizations are investing in Agentic AI for IT Operations today. That number represents real production deployments, not roadmap items, which means the monitoring gap AgentMon addresses is an active operational risk for more than half the enterprise market.
Competitive Positioning
Codenotary enters the AI agent observability space from a distinct angle. Most observability vendors are extending existing APM or tracing capabilities toward agentic workloads. Codenotary is coming from the supply chain integrity and trust direction, which gives it a different security posture. The combination of notarization, continuous verification, and now behavioral monitoring creates a lineage-aware approach to agent governance that APM-first vendors don’t naturally offer.
The competitive risk is that larger platform vendors, including hyperscalers offering managed agent runtimes, will build monitoring and remediation capabilities directly into their orchestration layers. Codenotary’s defensible position is that platform-native monitoring is inherently conflicted: the same vendor running your agents has a commercial incentive to minimize reported risk. An independent trust layer, with cryptographic proof of what agents did and when, has durable value precisely because it sits outside that relationship.
Looking Ahead
Short-Term: From Visibility to Governance
The immediate market test for AgentMon is whether organizations with live multi-agent deployments will layer in purpose-built observability, or rely on existing tooling that wasn’t designed for autonomous systems. Given that the organizations deploying agents most aggressively are also the ones under the most regulatory pressure, particularly in defense, government, and financial services, the demand signal looks credible. We expect Codenotary to announce integrations with the major agentic frameworks over the next 12 months, which will be a prerequisite for broad adoption.
Medium-Term: Autonomous Remediation as Standard Practice
AgentX arrives at a time when the market is moving faster than most security vendors acknowledge. As infrastructure fleets grow and AI-generated code increases vulnerability surface area, manual remediation simply cannot keep pace. The organizations that adopt autonomous remediation tooling in the next 18–24 months will develop the institutional experience required to govern it safely. Those that wait will face the choice of accepting a growing backlog or implementing remediation automation under pressure, which is the worst possible time to learn its failure modes.
Codenotary’s growth in defense and government provides a proof point that even the most risk-averse buyers are moving in this direction. The commercial enterprise market, where risk tolerance is higher and deployment cycles are faster, should follow. The broader question is whether Codenotary builds the ecosystem relationships required to become a platform of record for AI infrastructure security, or remains a specialized point solution. That outcome will depend heavily on how aggressively they pursue integrations with the cloud providers and agentic frameworks where enterprise AI workloads are actually running.
ServiceNow Q1 2026 Results: AI Platform Growth Accelerates
Cyera Acquires Ryft to Build Agentic AI Security Platform
Liferay CMS: Headless CMS Built for Enterprise Governance
ServiceNow Acquires Armis: AI Cyber Defense Platform Analysis
Qdrant Cloud Adds GPU Indexing, Multi-AZ, and Audit Logging
