Abnormal AI company profile: valuation, funding & revenue

This Abnormal AI company profile covers the company's valuation, revenue, funding rounds, investors and history. Abnormal AI works in cybersecurity, with a focus on AI-based email and behaviour security. It was founded in 2018 and is headquartered in San Francisco, California, United States. Its latest valuation on record is US$5.1 billion (August 2024). Every figure links to a public source, and anything the company hasn't disclosed is marked as such.

Revenue

More than US$200 million — Annual recurring revenue (company-reported), as of August 2024. Source: Silicon UK (company press release).

Funding history

Abnormal AI timeline

Key dates from the sources on this page, oldest first. Metric snapshots are listed after the funding events.

• 2018 — Abnormal AI is founded in San Francisco by Evan Reiser (CEO) and Sanjay Jeyakumar.
• 2022 — Series C: See source, valuing the company at About US$4B. Investors: Insight Partners (lead).
• Aug 2024 — Series D: US$250M, valuing the company at US$5.1B. Investors: Wellington Management (lead); Greylock, Menlo Ventures, Insight Partners, CrowdStrike Falcon Fund.
• August 2024 — Total funding: About US$546 million.

Investors

Overview

Abnormal AI, which operated for most of its history as Abnormal Security, uses artificial intelligence to stop email attacks and other threats that rely on deceiving people. Instead of looking only for known malicious links or attachments, its system learns the normal behaviour of every employee, vendor and application in an organisation, including who they usually email, how they write, when they log in and how money normally moves, and flags messages and activity that deviate from those patterns. This approach is particularly effective against business email compromise, vendor fraud and phishing messages written by AI that contain no obvious malicious content.

Evan Reiser, the CEO, and Sanjay Jeyakumar, the chief technology officer, founded the company in San Francisco in 2018. Both previously worked on machine learning for advertising at Twitter, where they built systems that modelled user behaviour at enormous scale. They applied similar techniques to security, betting that understanding normal behaviour would catch attacks that rule-based tools miss.

Products

• Inbound email security: detects phishing, business email compromise, invoice and vendor fraud, and account takeover attempts in Microsoft 365 and Google Workspace.
• Account takeover protection: spots compromised accounts across email and cloud applications based on unusual activity.
• Security posture management: finds risky configurations in cloud applications.
• AI security agents: automated assistants that triage user-reported emails, train employees and investigate incidents.
• Cross-platform coverage: extending behaviour models to collaboration tools such as Slack, Teams and Zoom, and to other SaaS applications.

How behavioural AI catches attacks

A typical business email compromise attack might come from a lookalike domain or a genuinely compromised supplier account, asking the finance team to change bank details for an upcoming payment. The email contains no malware and no suspicious link, so traditional filters let it through. Abnormal's models notice that the sender rarely emails this person, that the tone differs from previous messages, that the request involves payment details and that the domain was registered recently. Combining these signals, it flags or removes the message. Because Abnormal connects through cloud APIs rather than rerouting mail, it can be deployed in minutes and can remove messages even after delivery.

Funding

Abnormal raised early funding from Greylock Partners and Menlo Ventures, then Insight Partners. A 2022 Series C valued it at about US$4 billion. In August 2024 it raised a US$250 million Series D at a US$5.1 billion valuation, led by Wellington Management with participation from Greylock, Menlo Ventures, Insight Partners and CrowdStrike's Falcon Fund. Total funding reached about US$546 million. The company said its annual recurring revenue had passed US$200 million, with growth of more than 100% year over year at the time. We haven't found a newer priced round, so the data above uses the 2024 figures.

Competition

Abnormal competes with email security vendors such as Proofpoint and Mimecast, which have long dominated the secure email gateway market; with Microsoft's built-in Defender for Office 365; and with newer API-based players such as Sublime Security and Material Security. Microsoft is both a partner and a competitor, since most Abnormal customers use Microsoft 365. Abnormal's advantages are its behavioural AI, fast deployment and strong results against social engineering.

Abnormal AI at a glance

Abnormal AI (Abnormal Security) is a San Francisco cybersecurity company founded in 2018 by Evan Reiser and Sanjay Jeyakumar. It uses behavioural AI to stop email attacks, fraud and account takeovers. It raised a US$250 million Series D at a US$5.1 billion valuation in August 2024, led by Wellington Management, and reported ARR above US$200 million.

Timeline

• 2018: founded in San Francisco.
• 2020: launches its API-based email security product.
• 2022: Series C values it at about US$4 billion.
• August 2024: raises a US$250 million Series D at US$5.1 billion; ARR passes US$200 million.
• 2025: rebrands as Abnormal AI and expands AI security agents.

Why email attacks are so costly

Email remains the most common starting point for cyberattacks. The FBI's Internet Crime Complaint Center has consistently reported business email compromise as one of the costliest forms of cybercrime, with billions of dollars in reported losses each year. Generative AI makes the problem worse by letting attackers write convincing, error-free messages in any language and impersonate executives and suppliers at scale. That has increased demand for tools that judge context and behaviour rather than spotting obvious red flags.

Possible IPO

With ARR above US$200 million in 2024 and fast growth, Abnormal has been frequently mentioned as a candidate for a future IPO. Its CEO has spoken publicly about building a durable public company. Timing will depend on growth, profitability and market conditions for cybersecurity listings.

Questions for security teams

• How many attacks is your current email security missing, and of what types?
• How does the tool integrate with Microsoft 365 or Google Workspace?
• How are false positives handled, and how much analyst time does triage take?
• What data is accessed, and how is it protected?

Who uses Abnormal

Abnormal's customers include large enterprises in finance, healthcare, retail, manufacturing and technology, as well as public sector bodies and mid-sized organisations. Finance and accounts payable teams are particularly important beneficiaries, because vendor invoice fraud and payment redirection scams target them directly. Many customers keep Microsoft's built-in protection and add Abnormal as a second layer that focuses on the sophisticated attacks that get through.

AI on both sides

Attackers now use AI to write convincing emails, clone voices and research targets, while defenders use AI to detect anomalies and automate responses. Security leaders increasingly describe this as an arms race. Abnormal's bet is that behaviour-based AI, trained on how each organisation normally communicates, will remain effective even as attackers' messages become indistinguishable from legitimate ones in tone and grammar.

Latest developments (October 2026)

As of October 2026 we found no publicly announced funding round for Abnormal AI after its US$250 million Series D in August 2024 at US$5.1 billion, led by Wellington Management. Coverage at the time described the company as setting its sights on an eventual IPO, and later reports pointed to finance hires consistent with that preparation, but no listing has been announced. The email-security market around it has continued to shift toward AI-driven detection of business email compromise and vendor fraud, where Abnormal built its reputation, while large platforms such as Microsoft keep improving built-in protection. The 2024 valuation remains the latest sourced figure.

What to watch

Watch for a new funding round or IPO filing, ARR updates, adoption of its AI agents and competition with Microsoft and Proofpoint.

Takeaways

Organisations can add behavioural detection to catch fraud that filters miss. Founders can see how expertise from adtech machine learning transferred to security. Investors should watch growth relative to its 2024 valuation and IPO readiness.

Sources

Every figure on this page comes from the public sources below. Where sources disagreed, we used the company's own announcement or a primary news report. Private-company data changes; check the date beside each figure.

1. Crunchbase News
2. Silicon UK (company press release)