Abnormal AI announced an expansion of its Behavioral Security Platform with the introduction of Identity Threat Protection, AI Governance and Infiltration Prevention, extending the company's behavioral AI capabilities beyond email security to enterprise identities, AI systems and employee onboarding.

According to the company, the new products build on the behavioral AI technology it has used for nearly a decade to learn normal activity for each identity and detect behavior that may indicate malicious activity. Abnormal said the expanded platform now applies that behavioral model across a broader enterprise attack surface.

"Our customers are facing key challenges today: attackers who blend in instead of breaking in, AI adoption outpacing governance, and nation-state actors engineering fake identities to become insiders," said Evan Reiser, CEO and co-founder of Abnormal AI. "Every one of those risks comes down to understanding identity and behavior. That's the lens that Abnormal brings to these problems."

The company said modern attacks increasingly rely on trusted identities and authenticated sessions rather than stolen credentials alone. It also noted that organizations are managing growing numbers of non-human identities such as service accounts, OAuth applications and AI agents while AI adoption continues to outpace governance efforts.

Identity Threat Protection is designed to identify weaknesses such as accounts without multifactor authentication and overprivileged service accounts, ranking them by likely attack susceptibility. The product also includes a Threat Library that maps attack techniques to a customer's environment and correlates identity, SaaS and email signals to help detect compromised accounts after authentication. It also challenges suspicious password and MFA reset requests to help protect help desk operations.

AI Governance is intended to help organizations discover, score and govern AI tools, agents and chats across both sanctioned and unsanctioned environments. The product establishes behavioral baselines for AI tools and agents, identifies activity that deviates from expected behavior and automatically enforces customer-configured policies when risky activity is detected.

Infiltration Prevention focuses on identifying potentially fraudulent job candidates before enterprise access is granted. The product integrates with applicant tracking systems including Greenhouse and Workday and analyzes behavioral indicators such as VoIP burner numbers, VPN-masked locations and recurring threat intelligence signals to help security teams review suspicious applicants before accounts are provisioned. It also identifies broader campaigns and generates evidence briefs for investigation.

"The threats that matter most don't look like threats at all," Reiser said. "A trusted identity moves through a valid session like a real employee. A helpful AI agent codes its way into a sensitive database. A nation-state actor uses a synthetic persona to apply for a job. Each looks different, but each gives itself away the same way, through abnormal behavior."

Alongside the new products, Abnormal AI introduced the AI App Store, a new interface within the Abnormal Portal that allows security teams to browse, trial and activate the company's behavioral AI products with a single click.

Identity Threat Protection, AI Governance, Infiltration Prevention and the AI App Store are scheduled to become available Aug. 3, 2026.