The Rise of Shadow AI
The rapid democratization of generative AI has created a unique security nightmare for enterprise IT departments. As non-technical staff in finance, marketing, and legal teams begin to deploy their own custom AI agents and connectors, they are inadvertently creating massive, exposed security gaps. Pluto Security, which recently emerged from stealth with $42 million in total funding, is betting that traditional security software is simply not equipped to handle the fluid, autonomous nature of these new AI-driven workflows.
According to CEO and co-founder Shahar Bahat, the urgency of this problem is underscored by the speed at which new tools propagate. Within just two hours of a new AI tool hitting the market, it can become embedded in an enterprise environment. Pluto’s research indicates that 60% of detected risks originate from tools created by non-developers, with finance departments currently serving as the most significant source of vulnerabilities—often involving sensitive customer information being funneled through poorly secured, 'vibe-coded' applications.
Shifting Focus: From User Behavior to Agent Behavior
Pluto Security differentiates itself by shifting the paradigm from monitoring human user behavior to supervising agent behavior in real-time. The platform acts as an observability and control layer that keeps track of the skills, extensions, and MCP (Model Context Protocol) servers that power modern AI agents. By integrating with existing endpoint security and device management systems, Pluto deploys sensors that analyze prompts, commands, and permissions.
Perhaps most importantly, the platform offers active intervention capabilities. If an agent attempts to exceed its intended scope or access unauthorized permissions, Pluto can intervene during the session. The system can block the specific action or inject contextual instructions back into the agent’s workflow, forcing the AI to pause and request human confirmation before proceeding. This allows organizations to maintain security posture without stifling the productivity or experimentation of their workforce.
Community-Led Growth and Future Expansion
Pluto’s strategy for market penetration is distinct. Alongside its enterprise offering, the company launched 'Plutonium,' a free community-focused platform that allows security professionals to audit the risk profiles of public AI skills, connectors, and add-ons. This platform serves as both a public good and a lead generation engine, with Bahat reporting that 70% of their sales pipeline originates from inbound interest, bolstered by the community feedback loop provided by Plutonium users.
The company’s growth has been swift. Since beginning commercial sales in April, Pluto has already secured over 30 enterprise clients, including notable names like Oscar Health, Workato, Sixt, and AlphaSense, and has already surpassed $2 million in annual recurring revenue. The $42 million in funding—composed of an $8.5 million seed round and a $34 million Series A—will be used to scale the team, deepen security research efforts, and expand the company’s footprint in the U.S. market.
Why It Matters
- Visibility Gap: Enterprises currently lack tools to track the proliferation of third-party AI agents and shadow IT components.
- Proactive Defense: The ability to intervene during agent execution is a critical shift from passive logging to active threat mitigation.
- The 'Vibe-Coding' Risk: Non-technical employees are now building sophisticated automation, making them the primary target for security oversight rather than traditional software developers.
