Abnormal Email Security
abnormal.ai
- Web
- API
What Abnormal Email Security is
Abnormal Email Security is a cloud service that uses behavioral AI to detect business email compromise, vendor fraud, account takeover, and AI-generated attacks. Its Attune foundation model builds behavioral baselines for individual identities and identifies deviations. The product page describes API deployment in 60 seconds. Search & Respond supports hunting, triage, and remediation across Microsoft 365 and Google, with more than 30 search filters. Custom AI Models can be created from email examples and natural-language descriptions and tested against historical messages. Detection 360 investigates submitted missed attacks or false positives, explains outcomes, and updates detection models. Reported-phishing analysis and remediation are automated, while Email Productivity filters newsletters, promotions, and other bulk messages. Behavioral protection also covers Slack, Microsoft Teams, and Zoom. Listed integrations include Microsoft Sentinel, Splunk Enterprise Security, IBM QRadar, Okta, Microsoft Entra ID, and ServiceNow. Pricing is available on request.
Who it suits
Abnormal describes this service for CISOs, security leaders, and SOC teams. It suits organizations using Microsoft 365 or Google that need behavioral email and messaging protection.
Good to know, and what to watch
- Behavioral baselines for individual identities
- API deployment described in 60 seconds
- Search across Microsoft 365 and Google
- Custom detection models
- Protection for Slack, Teams, and Zoom
- Pricing is available on request
- Cloud deployment only
- Per-user mailbox pricing varies by features
Plans
Every plan and priceAbnormal Email Security
Pricing on request
The verdict
Abnormal Email Security combines behavioral detection with investigation, custom models, phishing analysis, and graymail filtering. Teams should request pricing and confirm the integrations and platforms needed for their environment.
What the maker publishes
- Company
- Abnormal says it was founded in 2018; an official launch announcement identifies San Francisco, California as its base. source
- Custom models
- Custom AI Models lets users create detection models from email examples and natural-language descriptions, then test them against historical messages. source
- Deployment
- The product page says deployment is via API in 60 seconds. source
- Detection
- Abnormal says its Attune foundation model builds per-identity behavioral baselines and detects attacks by deviation from those baselines. source
- Email productivity
- Email Productivity filters graymail such as newsletters, promotions, and bulk messages. source
- Founded
- 2018 source
- Headquarters
- San Francisco, CA source
- Integrations
- Abnormal's integrations directory lists Microsoft Sentinel, Splunk Enterprise Security, IBM QRadar, Okta, Microsoft Entra ID, and ServiceNow among supported integrations. source
- Intended users
- Abnormal describes its solutions as purpose-built for CISOs, security leaders, and SOC teams. source
- Messaging security
- The product page describes behavioral AI protection for Slack, Microsoft Teams, and Zoom. source
- Operations
- Detection 360 investigates submitted missed attacks or false positives, explains outcomes, and updates detection models. source
- Privacy
- Abnormal says its customer Data Protection Addendum addresses privacy frameworks including GDPR and CCPA. source
- Purpose
- Cloud Email Security uses behavioral AI to detect email attacks, including business email compromise, vendor fraud, account takeover, and AI-generated attacks. source
- Reported phishing
- AI Security Mailbox automates analysis of user-reported phishing, user responses, and threat remediation. source
- Search and response
- Search & Respond supports hunting, triage, and remediation across Microsoft 365 and Google, with more than 30 search filters. source
- Security and compliance
- Abnormal says it maintains SOC 2 certification and is certified to ISO/IEC 27001:2022, ISO/IEC 27701:2019, and ISO/IEC 42001:2023. source
Checked against the maker's pages on 3 Oct 2026.