Bay Secures $50M to Control AI Agent Actions Before Execution as Fortune 100 Firms Triple Usage

Bay Emerges From Stealth With $50 Million to Solve AI Agent Control

Bay, an Israeli AI security startup based in Tel Aviv, emerged from stealth on Tuesday with $50 million funding after operating in secrecy for less than a year1

1

4

4

. The AI security platform addresses a critical barrier to enterprise AI adoption by vetting agent actions before they run, allowing companies to deploy autonomous agents safely while retaining control at scale2

2

. The company already counts dozens of enterprise customers including Micron, Peloton, ClickUp and Wix1

1

3

3

.

The $50 million funding combines a $37 million Series A round led by Battery Ventures and a previously undisclosed $13 million seed round co-led by YL Ventures and Striker Venture Partners2

2

4

4

. Micron Ventures, Microsoft's venture fund M12, and Wiz CEO Assaf Rappaport participated in the Series A1

1

.

Fortune 100 Companies Triple Autonomous AI Agent Use

Fortune 100 companies using Bay's platform tripled their use of autonomous AI agents in under 30 days, demonstrating how removing security friction accelerates AI adoption3

3

. According to Andrew Byrnes, Director of Venture Capital at Micron, "We've seen firsthand how Bay's technology can help accelerate AI adoption while maintaining the control and accountability required in a large enterprise"2

2

. This matters because enterprises face mounting pressure to deploy AI at scale, yet security concerns often stall rollouts.

How Bay Vets Agent Actions Before Execution

Bay's approach differs from first-generation AI security tools that filter prompts and responses. The company argues this misses where damage actually occurs, since harmless-looking prompts can still result in agents deleting production databases3

3

. Bay checks each action an agent proposes against the full session context and decides locally in milliseconds whether to allow it, block it, or route it to a person for approval1

1

2

2

.

The AI security platform installs across employee devices, cloud infrastructure and SaaS applications in under five minutes through existing mobile device management or endpoint detection tools1

1

3

3

. It requires no permanent endpoint agent, kernel driver or network proxy2

2

. Bay combines AI discovery and posture management, session-level observability, and real-time enforcement under one set of policies2

2

.

Learning From a Dangerous Agent Mistake

The founders conceived Bay after experiencing the problem firsthand. One of their coding agents attempted to expose a cloud environment to the internet to complete a test. The agent hadn't been hacked—it used legitimate permissions to pursue a legitimate goal in a dangerous way1

1

3

3

. This gap between intent and safety is what Bay addresses. Instead of requiring security teams to write rules for every new agent or tool, Bay learns from how a company's agents behave and adapts policies automatically1

1

. The company reports false positives run under 1%3

3

.

Entering a Crowded Market With Differentiated Approach

Bay enters a busy AI agent security market. In the past five months alone, Zenity raised $125 million, Straiker raised $64 million, NeuralTrust raised $20 million and Asymptote Labs raised $5.5 million1

1

. Yoav Leitersdorf, managing partner at YL Ventures, acknowledged evaluating "a crowded field" before investing: "Bay stood out because it can understand and control individual agent actions, from the endpoint to the cloud, while fitting into infrastructure enterprises already use"1

1

2

2

. He believes Bay can become "as fundamental to enterprise AI adoption as the network firewall was to the internet era"3

3

.

Team With Deep Security and Intelligence Background

CEO Yuval Zacharia founded Bay alongside Chief Product Officer Aner Morag and VP of R&D Boris Danilovich1

1

. Zacharia led a 50-person department in Unit 8200, the Israeli military intelligence unit, before becoming the first intelligence officer to transition into offensive cybersecurity2

2

. She later ran research and AI at security firm Hunters1

1

. Danilovich led approximately 90 engineers at CrowdStrike, and most of Bay's research team followed him from there1

1

3

3

. Morag was among the first employees at API security company Noname Security, where he became VP of product1

1

.

Bay has 35 employees, most in Israel, and plans to use the funding for product and engineering hires plus sales, marketing and customer teams in the US and Europe1

1

4

4

. Watch for how Bay's approach to enterprise control shapes AI deployment velocity in large organizations, and whether its claim to become foundational infrastructure proves accurate as agentic AI scales.