The Pulse
Metano exits stealth with $12 million for AI agent security
Metano has emerged from stealth with $12 million in seed funding and a generally available AI agent security platform. The company says its software discovers agents, assesses their behavior and applies security controls across enterprise environments.

Metano has emerged from stealth with $12 million in seed funding and a generally available platform for securing enterprise AI agents. Sierra Ventures led the seed round, joined by Correlation Ventures, High Alpha and 360 ONE, the company said in its November 19 announcement.
The San Jose company offers security and governance software intended to discover agents, assess their behavior and enforce controls across endpoints, cloud infrastructure and software-as-a-service environments. It also offers a free evaluation edition with a limited scanning period.
Finding agents and their access
Metano says the product discovers both approved and unapproved AI agents, along with MCP servers, skills and software packages. It also identifies their access privileges and potential security exposure.
Its approach is aimed at agents that can use legitimate credentials and approved tools while changing their behavior in response to instructions, external content or operating conditions. The company argues that existing security products cover individual infrastructure layers but lack the visibility and behavioral context needed to follow agents across them.
The company's website says controls for each agent are generated from the specification written by its developer and then calibrated to observed behavior. The platform is intended to cover endpoints as well as homegrown, cloud and SaaS agents.
Following a session
The company says its intelligence layer, Hive, connects an agent's tasks, identities, tools, data access and actions in a shared context graph. It analyzes activity across a session to identify suspicious behavior and assess whether actions match the agent's intended purpose.
Metano also says its controls operate where agents run, including on activity that may bypass traditional network gateways. Centrally managed policies are intended to help security teams detect and stop dangerous actions as they occur.
The company says its sensor's initial pass identifies agents, MCP servers and skills on endpoints and recommends whether to approve, restrict or block them. Runtime guardrails fitted to an organization's usage follow after about a week of observation, according to its website.
Team and research
Metano was co-founded by CEO Neelima Rustagi and CTO Ratnesh Pandey. The company draws on security experience from ServiceNow, Zscaler, Palo Alto Networks, FireEye and Microsoft.
Its dedicated research team, MARS Labs, investigates agent security threats, including malicious skills, compromised software dependencies and vulnerabilities in MCP connections. Metano says that research informs the platform's security capabilities.
Availability
The free Eval Edition offers 30 days of real-time agent scanning, intended to identify agent activity, risky configurations and potential exposure in an organization's environment. The company says security teams can begin using the platform in less than 30 minutes.
Its agent detections map to MITRE ATLAS and the OWASP Top 10 for Agentic Applications. Metano has completed a SOC 2 Type 2 examination, holds ISO 27001:2022 certification and is listed on the Okta integration networks.
Sources
- Original source Metano · Metano emerges from stealth metano.ai
Other pages read
- Metano · page 2 metano.ai
Related stories
- Researcher says CreditNirvana flaws exposed 6.6 million accounts
- Sakana AI forms security research group focused on AI defense
- ToolPulse researchers reconstruct AI agent workflows via cache traces
- Microsoft urges tighter security controls as frontier AI advances
- Reco reports eight Microsoft 365 takeovers, with AI role uncertain