Bay Raises $50 Million to Control What Enterprise AI Agents Can Do

Tel Aviv-based Bay has emerged from stealth with $50 million funding (comprising a $37 million Series A round led by Battery Ventures and a previously undisclosed $13 million seed funding led by YL Ventures and Striker Venture Partners). Bay was founded in February 2026 by Yuval Zacharia (CEO), Boris Danilovich (VP of R&D), and Aner Morag (CPO). All three are alumni of the Israeli Military Intelligence (IDF Unit 8200).

In business, companies cannot ignore the economic advantage offered by AI agents, but nor do they always know what any individual agent might do. Agentic AI behavior is decided at runtime by each agent acting on its user’s behalf and with its user’s permissions.

Bay is designed to allow companies to install and employ AI agents confident that they will not run amok with unintended actions. It provides a single control plane that discovers, understands and controls agents wherever they run.

When installed, Bay automatically builds a full inventory of the AI agents, coding assistants, MCP servers, skills, extensions and local models in use, and maps how they connect, including risky combinations of tools. It follows each agent session from the first prompt to the final action. Its Enforcer component evaluates every proposed action in the context of the full session and decides in milliseconds whether to allow the action, refer it to the user, redirect it based on policy, or simply deny it.

The policy guiding Enforcer’s actions is automatically built and maintained by Bay’s embedded AI system (Lighthouse) based on the organization’s own activity history.

By focusing on the agents’ actions, Bay protects against dangerous agent activity regardless of whether the cause is malicious human-directed prompt injection or organic agent purpose drift.

“Before we even decided what to build, we agreed on three non-negotiables,” Bay’s CEO and co-founder Yuval Zacharia told SecurityWeek: “effortless deployment, immediate time to value, and a clear path to operationalization.” 

These principles became the design foundation for Bay. “The first drove us to build a flexible, agentless architecture that can be deployed across an entire fleet in under five minutes,” she continued. “The second led to patented technology that retrospectively analyzes 30 days of historical logs, allowing customers to see what Bay would have delivered had it been running in their environment. This makes it possible to demonstrate business value almost immediately and complete a proof of value in less than 14 days.”

The third, she concluded, “shaped how we approach day-to-day operations, with clear ownership, AI-driven policy creation, and a product designed to become part of how teams actually work.”

The funding will be used to further develop the product and expand the firm’s engineering team, and to invest in go-to-market plans across the US and Europe, “Longer term,” explains the firm, “we want Bay to be the core control layer for enterprises putting AI agents to work at scale. As agents take on longer-running, more autonomous work, Bay’s job is to make sure organizations can trust them with it, and to keep expanding what agents can safely do on their own.”