Abnormal AI

Behavioral AI security platform protecting email, identity, and AI from phishing, fraud, and insider threats.

Abnormal AI models normal communication behavior across an organization, then flags messages that deviate from it. Its behavioral approach spans email, identity, and AI surfaces, and it targets novel attacks that evade traditional security tools, including business email compromise, phishing, and account takeover, while detecting compromised accounts and malicious insiders. Its cloud-native API architecture deploys in minutes with no agents or MX changes and integrates with Microsoft and Google email platforms.

Category

Social Engineering & Human Risk

Founded

2018

Funding

$534M raised (Series D, 2024)

Website

https://www.abnormal.ai

Seen at Black Hat USA 2026 (Lobby Lounge sponsor).

Outbound clicks are tokenized for directory analytics.