Offroad, a startup developing artificial intelligence tools for identity security, emerged from stealth Wednesday with $7 million in seed funding led by Ibex Investors and Skywell Capital.
The company says it is building an “agentic identity security team” for enterprises, using AI agents to investigate and remediate risks across human users, machine identities and AI agents.
Offroad said identity security has expanded beyond determining who has access to systems. Companies now must understand why access exists, whether it is still justified, how it is being used, who owns the decision and what could break if access is changed.
That context is often scattered across identity providers, human resources systems, software applications, cloud platforms, security tools, tickets, logs and business owners. Offroad says its agents gather that information, identify real-time identity threats and broader posture risks, and resolve them either by acting directly when safe or involving the appropriate people with the necessary context.
The company cited OAuth applications as one example of a growing identity risk. Such applications can receive broad and lasting access to systems including Google Workspace and GitHub, sometimes covering files, email, code, calendars and other sensitive business data.
Offroad said it audited 2,890 public OAuth applications listed in the Google Workspace Marketplace and GitHub Marketplace as of May 2026. It found that roughly one in three apps, representing more than 1.85 billion installs, showed serious structural security concerns that a security analyst would likely reject during manual review.
The company is also launching ohauth.ai, a free OAuth security catalog designed to help teams review app permissions, security concerns and governance risks.
“Identity is no longer just a workforce access problem,” said Offroad co-founder and CEO Dan Bendler. “Enterprises now operate across a constantly changing mix of human users, machine identities and AI agents.”
Offroad was founded in 2025 by Bendler, who previously founded two AI startups that raised more than $45 million combined, and CTO Philip Shteyn, a former Unit 8200 captain who helped build Palo Alto Networks’ Cortex platform.
“Most identity systems were designed around assumptions that no longer hold,” Shteyn said. “AI agents operate across systems at all hours and at a scale humans never could, which makes traditional behavioral baselines far less reliable.”
Sean Mullins, vice president of enterprise infrastructure and CISO, said Offroad’s agents are designed to reduce the burden on security teams rather than add more alerts.
“Rather than adding more alerts to an already noisy environment, they gather context, work through the risk, and hand off something actionable,” he said.
Adi Dangot Zukovsky, a partner at Ibex Investors, said companies increasingly rely on autonomous identities with broad permissions.
“Security teams do not need another tool that provides findings,” she said. “They need operational leverage, not more interfaces.”