Ex-Run:ai and Orca Security leaders raise $8M in Seed funding to build AI-native identity platform
Fabrix says its explainable AI agents will transform access governance at scale.
Identity and access management, the decades-old challenge of controlling who has access to what within corporate systems, has long been seen as a cumbersome but unavoidable task. Now, a Tel Aviv-based startup, Fabrix Security, is betting that artificial intelligence can finally resolve one of cybersecurity’s most persistent bottlenecks.
Full list of Israeli high-tech funding rounds in 2025
The company emerged from stealth on Tuesday with $8 million in Seed funding led by Norwest and toDay Ventures, with support from Jibe Ventures. Additional backers include Shmil Levy, co-founder of Sequoia Israel; Yotam Segev and Tamar Bar-Ilan, founders of Cyera; Ronen Dar and Omri Geller, founders of Run:AI (acquired by NVIDIA); Ofer Ben-Noon and Ohad Bobrov, founders of Talon (acquired by Palo Alto Networks); Yoni Broyda, founder of Alooma (acquired by Google); and Michael Dolinsky, co-founder of Ermetic (acquired by Tenable) and Aorato (acquired by Microsoft).
Founded by CEO Raz Rotenberg, an AI expert and a founding engineer and R&D lead for Run.ai, which was acquired last year by NVIDIA, and CTO Ofir Yakovian, the former Technical Lead for Orca Security and Microsoft Entra, Fabrix already counts paying customers and partnerships with leading U.S. financial institutions.
Fabrix’s pitch is simple: traditional identity systems, which rely on manual processes and rules-based reviews, cannot keep pace with the explosive growth of both human and non-human identities, such as bots, service accounts, and AI agents. As enterprises add more cloud applications and AI-driven workflows, the attack surface expands. “At this scale, traditional IAM systems can’t achieve their main objectives, enforcing least privilege and reducing the attack surface,” said Rotenberg, Fabrix’s CEO.
Related articles:
Instead, Fabrix offers what it calls “AI-native IAM.” Its system uses connectors to pull data from across an organization’s networks, building what it describes as an “identity fabric graph.” AI agents then analyze that graph, making access decisions in real time and adapting them dynamically to shifting business contexts. Unlike static tools, Fabrix emphasizes explainable, auditable reasoning, designed to reassure risk managers who have been wary of AI’s opacity.
