Brava Security raises $15M to simulate attacks and close AI detection gaps

Tel Aviv-based Brava Security has emerged from stealth with $15 million in seed funding, targeting a critical gap in cybersecurity: the inability of AI-driven defenses to reliably identify attacks before they occur. The round was led by F2 Venture Capital and Pitango VC, with additional support from angel investors, as enterprises struggle with the limitations of current detection tools.

The company’s platform uses AI to simulate attacker behavior, generating attack telemetry that reveals blind spots in existing security infrastructure—particularly in SIEM systems, which often fail to detect threats due to a lack of environmental context. Unlike traditional approaches focused solely on detection rules, Brava’s method validates how attacks would materialize across an organization’s log sources, ensuring defenses are tested against realistic scenarios. The founders—both with backgrounds in Israeli intelligence—argue that security tools must first confirm they can detect threats before automation can effectively respond.

The timing is significant. As AI agents become more prevalent in enterprise environments, often operating with unmonitored identities, startups like Brava are addressing a key vulnerability: the assumption that security tools will catch exploits before they cause damage. Earlier this year, Rig Security raised $12 million to monitor AI agents as they move through systems, while Brava takes a different approach by validating detection capabilities upfront. If an AI agent—or a human attacker—exploits a weakness, how can organizations be certain their tools will recognize it?

Brava’s solution integrates with over 70 data sources and destinations without requiring infrastructure changes, and claims to provide initial insights within 10–15 minutes of deployment—a notable advantage in a space where security tools often demand lengthy implementation cycles.

The funding will support further product development and expansion, with Brava maintaining an R&D team in Israel and a commercial presence in the U.S. The startup’s emphasis on simulation over reactive analysis reflects a broader industry shift: moving from detection-centric models to prevention through predictive modeling. This approach is especially relevant for enterprises deploying AI agents, where the attack surface is expanding faster than traditional defenses can adapt.

The question remains whether Brava’s simulation-first strategy can scale beyond early adopters—and whether the market will prioritize startups that address "unknown unknowns" in cybersecurity. If successful, the company could redefine how security teams validate their defenses, shifting the focus from catching threats after they occur to ensuring they’re detected in the first place.

Sources: securityweek.com