Insygna launches free AI agent security report card

Insygna on August 10 launched a free public service that scores AI agents for security risks before they get access to company systems. The move targets a fast-growing market where many enterprises lack controls and where rogue agents can leak data, trigger costs, or act with excessive permissions.

Why it matters: - Companies are deploying AI agents faster than they can verify them. - Insygna's new service is designed to give executives an independent security check before an agent touches company or client systems. - The goal is to reduce the risk of data leaks, hidden-instruction attacks, unauthorized actions and cost overruns. - The concern is higher for companies using AI agents they did not build themselves.

What happened: - Insygna launched the Insygna Agent Report Card, a free public service for independently testing AI agents. - The service went live on August 10, 2026. - A company creates a free account, connects its AI agent repository, and receives a score for each agent. - Each report also includes a shareable Insygna Verified™ badge and a full list of findings with version control. - The report card is available now at the Insygna Agent Report Card.

The details: - The score works like a credit score for an agent and runs from 0 to 100. - The score is weighted across six security dimensions: Secret Exposure, Dependency Vulnerabilities, Code Security, Container Hardening, LLM Security measured against the OWASP LLM Top 10, and Image Security. - Every finding is identified down to the file, the line and the fix. - Insygna says an executive can require an Insygna report card before any agent goes live and can hold the deployer accountable for the score. - The median AI agent Insygna has tested so far scores below 50 out of 100. - Roughly six in ten tested agents score below 50. - Insygna said millions of agents could represent a broad industry risk if those results hold at scale. - The Agent Report Card is part of Insygna Agentic Workforce Management™, which also includes ongoing monitoring, continuous reporting and management of AI agents as an extended workforce. - The platform connects with tools including Microsoft Teams, Slack, Copilot, Claude and ChatGPT. - AI agents built on Oracle, Salesforce, Workday, ServiceNow and other platforms can be connected to Insygna through GitHub. - Insygna helps companies manage AI agents using existing HR, Finance and Procurement processes. - The platform gives each agent a verifiable identity and a managed lifecycle from deployment to retirement. - Insygna says the system helps companies control agent costs, protect proprietary data and stay compliant while tracking which agents are running, what they can do and who is accountable. - Insygna is headquartered in Merrimack, New Hampshire, and has a development center in Vilnius, Lithuania. - The company won the HR Tech Europe 2026 Startup Competition and was shortlisted for a 2026 HR Tech Product of the Year award.

Between the lines: - The launch frames third-party verification as a trust layer for a market that is moving faster than internal governance. - CEO Michael Beygelman said the free service was mobilized because companies need a reputable third party to test AI agents in real time and report risk honestly. - Beygelman also said relying on the same platform that hosts an agent to grade that agent is a less defensible practice. - The pitch positions Insygna as an outside validator for agents built on major enterprise and AI platforms.

What's next: - Companies can begin using the free report card immediately. - Insygna is likely to use the service to funnel users into its broader monitoring and management platform. - As agent adoption grows, demand for independent testing and lifecycle controls is likely to increase.

The bottom line: - Insygna is betting that AI agents will need a security score before they get access to enterprise systems, and that third-party validation can become a new standard for trust.