Pluto Security, a startup focused on securing the use of artificial intelligence within enterprises, has raised $42 million in funding after emerging from stealth, at a time when companies are increasingly concerned about tools and applications created or used by employees without adequate security review.
Shahar Bahat, the chief executive officer and co-founder, says a new AI tool may appear within a company’s environment just two hours after its launch. According to Pluto’s data, about 60% of the risks it identifies among its customers are associated with tools created by non-developers using what is known as “vibe coding.” Finance departments lead, followed by marketing and legal affairs, while developers rank fourth.
Risks of Unreviewed Applications and Agents
Pluto’s tools detected an application created by an employee in the finance department of a Fortune 500 company to manage customer data and churn rates, but it was left exposed on the internet. Bahat says the proof-of-concept tests conducted by the company were not free of malicious elements, including extension packages and skills that contained active attacks during testing.
Pluto is betting that traditional security tools alone are not enough to monitor an environment in which the way applications are built and used is changing rapidly. Instead of focusing only on user behavior, the company wants to monitor the behavior of agents and the processes they execute moment by moment.
Monitoring Tools and Intervening During Execution
Pluto’s software monitors the AI tools employees use, in addition to the skills, extensions, MCP servers, and other components associated with them, as well as the applications and agents they create. The company says the system can identify dangerous components, detect suspicious agent activity, and enforce security policies without completely preventing employees from using or developing the tools.
The system connects to endpoint protection and management platforms, and can deploy sensors directly on devices or through those platforms. It also collects data including prompts, commands, ongoing processes, permissions, and previous activity logs, and uses lightweight models on endpoints to help assess what the agent is attempting to do.
One of its key advertised capabilities is intervening while an agent is operating. If, for example, it attempts to obtain permissions beyond what the user intended, the system can stop the action or add instructions to the session asking the agent to obtain confirmation.
Funding and Commercial Growth
The funding is split between an $8.5 million seed round led by TLV Partners, with participation from the founders of cybersecurity company Armis, and a $34 million Series A led by Jump Capital, with participation from Skinos, the venture capital firm affiliated with Check Point co-founder Shlomo Kramer, and other angel investors.
Pluto began selling its products in April and says it has exceeded $2 million in annual recurring revenue, with more than 30 enterprise customers, including Oscar Health, Workato, Sixt, and AlphaSense. The company has 40 employees in the United States and Israel, and plans to use the funding to develop the product, expand security research, and strengthen its presence in the United States.
Why Does This Matter?
The actual change here is the shift of the security monitoring point from employees’ accounts and traditional applications to the tools, agents, and components that may act on their behalf. However, the figures provided are based on the company’s statements, and the source does not provide independent details about the methodology for measuring risks or customer outcomes, nor does it precisely define the system’s intervention limits or false-positive rate. Therefore, the ability of this category to strike a balance between oversight and giving employees the freedom to use AI remains an open question.