Abnormal AI
Abnormal AI uses behavioural analysis to detect and stop email, identity, and AI-driven cyber attacks.
- Categories
- Email security, Identity threat detection and response, Data loss prevention, Security awareness and human risk
- Based in
- San Francisco, United States
- Founded
- 2018
- Ownership
- Private
- Certifications
- SOC 2, ISO 27001, ISO 27701, ISO 42001, FedRAMP, GovRAMP Moderate
About Abnormal AI
Abnormal AI sells a behavioural security platform that covers email, identity and AI. It is aimed at security leaders, SOC teams, IT and identity teams, and compliance and risk teams. The company says 5,000 organisations use it, including 30% of the Fortune 500. It also names sectors such as financial services, healthcare, technology, manufacturing, retail, public sector and education. The platform runs on Attune, a behavioural foundation model. Attune learns what normal activity looks like across identity, behaviour and content, then flags deviations. Abnormal says this needs no custom rules or manual tuning and adapts as attacks change. The Identity Trust Graph maps human, non-human and cloud identities. The platform ingests behavioural signals through API integrations and deploys in minutes without agents or MX changes.
Products
- AttuneA behavioural foundation model that learns normal activity and detects deviations across the platform.
- Inbound Email SecurityBlocks business email compromise, phishing and credential theft before messages are delivered.
- AI Security MailboxAn AI co-worker that supports email security operations.
- AI Phishing CoachTraining that adapts to the risk each user actually faces.
- Email ProductivityCuts graymail and unwanted messages so inboxes stay focused.
- Misdirected EmailStops accidental data leakage from email sent to the wrong address.