Orion Security

Orion Security is building what it calls “DLP beyond policies”: rather than relying on the manually authored, constantly maintained rule sets that make traditional data loss prevention brittle and noisy, its AI agents analyze the data itself, the systems it moves through, the people involved, and the surrounding business process to make contextual decisions about whether a data movement is actually risky.

Founded in 2024 by CEO Nitay Milner, a former product leader at Cisco-acquired Epsagon, and CTO Jonathan Kreiner, former application security lead at WalkMe, Orion is based in New York with an additional office in Tel Aviv. The company raised a $32 million Series A in February 2026 led by Norwest Venture Partners, with participation from IBM Ventures and existing investors PICO Venture Partners and Lama Partners. IBM published its own rationale for the strategic investment, a notable independent corporate-VC signal beyond the round itself.

Policy-free, contextual DLP is a meaningful rethink of one of security’s oldest and most maintenance-heavy categories, and the specific complaint it targets — brittle rules generating constant false positives — is real and well documented industry-wide. Several other data-security vendors are converging on similar AI-context approaches at the same time, though, and Orion hasn’t yet published named customers or independent test results, so its efficacy claims remain largely unproven in public.