A Tel Aviv cybersecurity startup, founded in 2025, has raised $12 million to tackle what its founders describe as the messiest frontier in enterprise security: telling the difference between a human clicking a mouse and an AI agent acting on that human's behalf.
Rig Security announced the seed round on September 29, 2026, co-led by Ten Eleven Ventures and Brightmind Partners. The CrowdStrike Falcon Fund and Ami Luttwak, co-founder and CTO of cloud security giant Wiz, participated as well. Calcalist reported the deal closed towards the end of 2025.
The company's pitch rests on a problem that has crept up on security teams faster than many expected. As businesses deploy more AI agents to handle routine tasks, those agents are logging in under employee credentials, making decisions, moving data and occasionally doing things their human overseers never intended. Traditional identity systems, built to manage people and their passwords, struggle to track what the software is doing once it has borrowed someone's access.
"Identity used to mean people's credentials," said Guy Kozliner, Rig's founder and CEO. "Today, the most active identities in many organizations are AI agents, and they are acting under human names."
Rig's platform layers identity security posture management with real-time threat detection, aiming to unify oversight across human users, machine identities and the growing population of AI agents. The underlying technology, which the company calls RICE—Rig Identity Correlation Engine—maps identities across cloud providers, on-premises systems and endpoints. Internal testing suggests the engine achieves better than 96 percent accuracy, though independent validation of the 96% accuracy is not publicly available.
The system also includes a lightweight endpoint sensor, identified by SecurityWeek as "Bifrost," that sits on devices and analyzes whether an action originated with a human or an agent using the same account. According to Rig, the sensor can block risky agent behavior inline, before it reaches cloud infrastructure.
Whether enterprises will trust an early-stage vendor to make those split-second blocking decisions remains to be seen. Rig said it already counts Fortune 200 customers across financial services, insurance, healthcare and technology, though it named only one publicly: New American Funding, a mortgage lender. Bill Harper, the company's AVP of Digital Identity, said in a statement that Rig helps his team see "which agents are operating [and] whose access they are using."
The startup's 20-person team carries notable pedigree. Kozliner worked on the CTO team at Wiz under Luttwak. CTO Nokky Goren (also referred to as Arthur Goren in some sources) was the first engineer at Axis Security, which Hewlett Packard Enterprise acquired in 2023. Head of Product Michal Haikov came from Unit 8200 and Flow Security, which CrowdStrike later bought.
Rig's emergence coincides with a flurry of funding for startups confronting AI agent risk. AIR raised $50 million across two seed rounds, according to TechCrunch, which also noted work by Astrix Security and Operant AI in adjacent areas. Above Security came out of stealth with $50 million for an agentic insider-risk platform, and Beacon Security raised $13 million for an AI agent data foundation aimed at cyber defense. (Note: some of these funding announcements reference future dates in the original draft, suggesting either placeholder timelines or information that may require verification.)
Reports of rogue agent behavior have filtered into public view. The Associated Press carried accounts of AI agents allegedly scanning UN statistics databases and U.S. government websites without clear authorization. Nvidia announced an "Open Agent Safety Platform" to address what it called the need to contain runaway agents, though the timing of that announcement in the original draft appears inconsistent with the current calendar.
"Rig goes further, connecting visibility, risk and real-time enforcement so security teams can actually act on identity risk," said Mark Hatfield, co-founder and partner at Ten Eleven Ventures.
The company said it already protects tens of millions of identity activities and has listings on AWS Marketplace and CrowdStrike Marketplace. It was a runner-up in a recent AWS and CrowdStrike Cybersecurity Startup Accelerator, a credential that likely helped open doors with enterprise buyers wary of unproven vendors.
Still, Rig faces the challenge every security startup confronts: convincing organizations to add yet another monitoring layer to an already crowded stack. The bet is that the chaos introduced by agentic AI will be urgent enough, and visible enough, that security teams will have no choice but to treat agent identities as a distinct problem requiring distinct tools.
