Pricing: Custom (quote-only, per-mailbox/year)
Best for: Research & Analysis
About
Abnormal AI (formerly Abnormal Security) uses behavioral AI to detect advanced phishing, business email compromise (BEC), and social engineering attacks that bypass traditional email security. It integrates with Microsoft 365 and Google Workspace to stop attacks that evade SEGs.
In-Depth Review
Abnormal AI (formerly Abnormal Security) addresses the biggest gap in enterprise email security: attacks that don't use malicious links or attachments — business email compromise (BEC), vendor fraud, and social engineering attacks that use clean domains, convincing language, and no traditional IOCs.
Why Traditional Email Security Fails
Secure Email Gateways (SEGs) like Proofpoint and Mimecast rely on known threat indicators — malicious URLs, suspicious domains, known malware signatures. BEC attacks use none of these. A CFO receives a convincing email from an attacker impersonating the CEO asking to wire money to a new vendor — the email comes from a clean domain, uses perfect English, and contains no malicious links. Traditional tools miss it.
Behavioral AI Baseline
Abnormal AI builds a behavioral baseline for every person in your organization — email communication patterns, typical senders, normal request types, writing style, business relationships. When an email deviates from baseline (an unfamiliar sender requesting financial action, an unusual wire transfer request, a vendor claiming a new bank account), Abnormal flags it regardless of technical indicators.
Identity Analysis
Abnormal analyzes the full identity context around every email: Is this sender's domain newly registered? Does the email header match the display name? Has this sender ever corresponded with this recipient before? Does the request pattern match previous interactions? Identity analysis catches impersonation attacks where the display name looks right but the underlying domain is wrong.
Attack Detection Types
Abnormal covers: business email compromise, phishing, vendor email compromise, payroll diversion, credential harvesting, account takeover (detecting compromised legitimate accounts), and supply chain fraud.
Remediation
When Abnormal detects an attack, it can automatically remediate — removing the email from all recipients' inboxes, notifying security teams, and providing attack analysis. The post-attack report explains exactly what happened, who was targeted, and what indicators triggered the detection.
Pricing
Abnormal AI pricing is quote-only and billed per-mailbox per-year (not a simple monthly rate) rather than published on a rate card — typical list pricing runs roughly $15-40/mailbox/year depending on deployment size and modules, often bundled with a flat platform fee. Minimum annual contracts typically start around $25,000-$50,000 for smaller deployments, with large enterprise deployments running into six figures annually.
Pricing
Custom (quote-only, per-mailbox/year)
Capabilities
Technical
- API Available
- Yes
- Languages
- English
Categories
Pros & Cons
Pros
- API available for developers
- Highly rated by users
Cons
- Limited to English
Related Chatbots
Claude
Claude is Anthropic's AI assistant, built for accuracy, nuance, and careful reasoning. It runs on the Claude 5 model fam...
Perplexity
Perplexity is an AI-powered search engine that answers questions with cited sources from the live web. Every response li...
DeepSeek
DeepSeek is a free AI chatbot from Chinese quantitative hedge fund High-Flyer that delivers frontier-level reasoning and...
Doubao
Doubao is ByteDance's AI chatbot and China's most-used AI assistant, serving over 155 million weekly active users. Power...
Explore More
Frequently Asked Questions
- Is Abnormal AI (formerly Abnormal Security) free to use?
- Abnormal AI (formerly Abnormal Security) is a paid tool, starting from Custom (quote-only, per-mailbox/year).
- What can Abnormal AI (formerly Abnormal Security) do?
- Abnormal AI (formerly Abnormal Security) supports bec-detection, phishing-prevention, behavioral-ai, email-security, account-takeover-detection, social-engineering-defense. Abnormal AI (formerly Abnormal Security) uses behavioral AI to detect advanced phishing, business email compromise (BEC), and social engineering attacks that bypass traditional email security. It inte
- Is Abnormal AI (formerly Abnormal Security) good for research & analysis?
- Yes, Abnormal AI (formerly Abnormal Security) is well-suited for research & analysis. Abnormal AI (formerly Abnormal Security) uses behavioral AI to detect advanced phishing, business email compromise (BEC), and social engineering attac
- Does Abnormal AI (formerly Abnormal Security) have an API?
- Yes, Abnormal AI (formerly Abnormal Security) has a public API available for developers.
- What languages does Abnormal AI (formerly Abnormal Security) support?
- Abnormal AI (formerly Abnormal Security) primarily supports English.
Know a tool we're missing? Submit it free →
Like what you see?
Get weekly chatbot news, reviews, and discoveries delivered to your inbox.