Bloom Security Raises $20 Million to Protect Corporate Data from Unknown AI Tools
Cybersecurity startup Bloom Security has raised $20 million in a seed round. The company is developing a platform to manage risks associated with autonomous AI agents and extensions running on employee computers.
The cybersecurity company Bloom Security has raised $20 million in a seed round led by Glilot Capital, with participation from Ten Eleven Ventures, Okta Ventures, and Runtime Ventures. Private investors also participated in the round, including the founders of cybersecurity companies Snyk, Demisto, Dig Security, and Talon.
Bloom Security was founded in 2025 by Itay Keren (CEO), Ofir Belsiano (CPO), and Itay Frishman (CTO), all veterans of the Navy, Mamram, and Unit 81. The founders were among the first employees at Dig Security and Demisto, both of which were acquired by Palo Alto Networks. The company currently employs 30 people in Israel, most of them from Dig Security and Palo Alto Networks.
In recent years, employee and developer computers have become complex software environments. Alongside the operating system and software installed by the organization, they run AI-based agents, MCP servers, extensions for browsers and development environments, automation tools, and code libraries. Browsers, development environments, and agents have themselves become platforms with app stores and tools for installing and managing code packages, so the number of programs running on a computer is growing faster than security teams can track them. Employees install some of the tools themselves, and others are added by AI tools. These tools can obtain permissions and access information and additional services without undergoing an orderly review process by the organization.
Bloom Security has developed a security platform for endpoints in the AI era. The platform is based on technology that runs autonomous AI agents, which analyze and map the programs, extensions, and code running on each computer, and examine what information and systems they access and how they operate alongside other tools. The risk assessment is also based on context: a program that is suitable for one employee may create a risk on another computer, depending on the user's role, the information they have access to, and the additional tools running in that same environment. The platform analyzes permissions, settings, and risks in the software supply chain and allows for blocking dangerous installations, enforcing security settings, and addressing risks without disrupting the user's work.
"Employee computers no longer include only software that the organization has chosen, tested, and decided is safe. They run agents, extensions, and code packages that connect to services and gain access to information, sometimes without the security teams knowing about it. We founded Bloom to enable organizations to understand what is actually running on every computer and to control risks, without delaying work and the adoption of AI tools," said Itay Keren, CEO of Bloom Security.
Kobi Samboursky, Managing Partner at Glilot Capital, added:
"The work environment of employee computers is changing rapidly, and existing security systems are failing to keep up. Agents, extensions, and code packages that run directly on computers create a new layer of risk that existing tools were not designed to handle."
