Pluto Security discloses $42 million for controls that interrupt risky AI-agent actions

Employees using AI to build workplace apps are creating security problems outside traditional development teams, according to Pluto Security. To expand its business addressing those risks, the startup emerged from stealth on October 13, 2026, disclosing about $42 million in seed and Series A funding. It says its software can interrupt risky AI-agent actions.

Jump Capital led the Series A, with Skinos, Ansa Capital and angel investors participating. The financing follows an $8.5 million seed round led by TLV Partners, which also drew investment from the founders of cybersecurity company Armis. The disclosed total covers both rounds, not a single $42 million investment.

The exact Series A amount remains unresolved: TechCrunch puts it at $34 million; PYMNTS, citing a LinkedIn post from Jump Capital, gives $33 million. Both put total funding at about $42 million.

Pluto plans to spend the capital on product development, expanded security research and a larger U.S. presence.

CEO and co-founder Shahar Bahat said Pluto began selling in April and has crossed $2 million in annual recurring revenue—the annualized value of recurring sales. She told TechCrunch it has more than 30 enterprise customers, including Oscar Health, Workato, Sixt and AlphaSense.

Finance teams, not developers, top Pluto’s risk list

Bahat told TechCrunch that about 60% of the risks Pluto detects across its customers are in tools built with AI by non-developers. Finance departments are the biggest source, followed by marketing and legal; developers rank fourth. That breakdown describes Pluto’s detections, not risk across all enterprises.

One example involved a finance employee at a Fortune 500 company who used AI to create an app for managing customer information and churn—tracking customers who leave. Bahat said Pluto found the app exposed to the internet.

A security layer that can interrupt an agent

Pluto’s software watches employee AI tools, their extensions and other connected components, and the apps and agents employees create. The company says it can identify risky components and enforce security policies without stopping employees from using or making AI tools.

• Deployment: Pluto connects to existing device-management and endpoint-security systems. It can install its sensors through those systems or directly on devices.
• Visibility: It gathers prompts, commands, running processes, permissions and users’ previous activity.
• Assessment: Its own models run on devices to help interpret what an agent is trying to do.

The intervention happens while an agent is working, Bahat said. If it tries to obtain permissions beyond what its user intended, Pluto can stop the action. Alternatively, the system can add instructions to the session that prompt the agent to ask for confirmation.

A free route to prospective customers

Pluto also has a free entry point: Plutonium, launched in April to check risks in publicly available AI skills, connectors and add-ons. Bahat said its community supplies feedback and prospective customers. Inbound interest accounts for 70% of Pluto’s sales pipeline, she said, without specifying how much comes directly from Plutonium.