Abnormal AI
About Abnormal AI
Abnormal AI is a behavioral AI security platform that protects Microsoft 365 and Google Workspace email against phishing, business email compromise, vendor fraud and account takeover. Founded in 2018 by Evan Reiser and Sanjay Jeyakumar and now based in Las Vegas, the company rebranded from Abnormal Security in April 2025 and moved its primary domain from abnormalsecurity.com to abnormal.ai. Deployment is API-only with no MX change, and detection runs on Attune, its behavioral foundation model released in March 2026. Email modules include Inbound Email Security, Account Takeover Protection, AI Security Mailbox for autonomous triage of user-reported phishing, AI Phishing Coach simulations and training, Posture Management for Microsoft 365, Messaging Security for Slack and Teams, Misdirected Email prevention and Email DLP Rules. In 2026 the platform expanded beyond email with Identity Threat Protection, AI Governance for shadow-AI discovery, Infiltration Prevention for fake job applicants and an AI Cloud Security preview. Abnormal holds FedRAMP Moderate and GovRAMP Moderate authorizations alongside SOC 2 Type II, ISO 27001 and ISO 42001, integrates with Microsoft Sentinel, Splunk, CrowdStrike, Okta and ServiceNow, and reports more than 4,500 customers including over 25 percent of the Fortune 500. Gartner named it a Leader in the December 2025 Magic Quadrant for Email Security, placing it furthest on the Completeness of Vision axis.
Why Notable
Gartner Magic Quadrant for Email Security Leader (Dec 2025, furthest on Completeness of Vision among 14 vendors); Forrester Wave Leader Q2 2025; Gartner Peer Insights Customers' Choice 2026 (4.8/5, 182 reviews); $250M Series D at $5.1B valuation led by Wellington Management (Aug 2024); FedRAMP Moderate and GovRAMP Moderate; rebranded from Abnormal Security to Abnormal AI in April 2025
