Startup Virtue AI Inc. today introduced AgentSuite, a software plaform designed to protect artificial intelligence agents from threats such as malicious MCP servers.
The platform’s debut comes less than a year after the company launched with $30 million in initial funding. It raised the capital from a consortium that included Intel Corp. Chief Executive Officer Lip-Bu Tan.
AI agents access the external applications they use to automate manual work through programs called MCP servers. An MCP server includes, among others, a so-called tool definition. That’s a machine-readable description of the application to which the MCP server is attached and the ways AI agents can use it.
Virtue AI says that AgentSuite checks tool definitions for malicious prompts. The software could, for example, detect if an MCP server instructs the AI agents that access it to disclose their training data. The platform also scans MCP servers’ code for known vulnerabilities.
AgentSuite includes a module called the Unified Agent Gateway that acts as an intermediary between AI agents and MCP servers. Tool use requests go through the module, which blocks them if they don’t comply with a company’s cybersecurity policies. Virtue AI says that Unified Agent Gateway adds less than 10 milliseconds of latency to prompt responses.
The platform can identify cybersecurity issues in not only the MCP servers that an AI agent uses but also the agent itself. According to Virtue AI, AgentSuite searches for weak points by simulating cyberattacks. It carries out those cyberattacks in more than 50 sandboxes designed to mimic the environments where AI agents operate. Some of the environments emulate a browser, while others include tools such as command line interfaces.
AgentSuite can prevent AI agent vulnerabilities from rolling out to production by integrating with a company’s CI/CD, or continuous integration and continuous deployment, system. Developers use CI/CD systems to automate the manual involved in releasing new software. AgentSuite flags vulnerabilities before the AI agent update that contains them is deployed.
Another task that the platform automates is spotting regulatory compliance issues. A bank, for example, could use it to detect when an AI agent generates investment advice in breach of financial industry rules. AgentSuite also spots more subtle issues. The software flags situations where an AI agent doesn’t carry out a task that it’s designed to perform because its guardrails are too strict.
“Agentic AI is becoming core enterprise infrastructure,” Virtue AI co-founders Bo Li, Dawn Song and Sanmi Koyejo wrote in a blog post today. “Securing it requires more than isolated controls or point solutions.”
Virtue AI disclosed that its platform has already been adopted by several major players in the AI market. The company’s installed base includes Nvidia Corp., Microsoft Corp., Google DeepMind, OpenAI Group PBC and others.
Image: Unsplash
Support our mission to keep content open and free by engaging with theCUBE community. Join theCUBE’s Alumni Trust Network, where technology leaders connect, share intelligence and create opportunities.
- 15M+ viewers of theCUBE videos, powering conversations across AI, cloud, cybersecurity and more
- 11.4k+ theCUBE alumni — Connect with more than 11,400 tech and business leaders shaping the future through a unique trusted-based network.
About News Media
Founded by tech visionaries John Furrier and Dave Vellante, News Media has built a dynamic ecosystem of industry-leading digital media brands that reach 15+ million elite tech professionals. Our new proprietary theCUBE AI Video Cloud is breaking ground in audience interaction, leveraging theCUBEai.com neural network to help technology companies make data-driven decisions and stay at the forefront of industry conversations.
