Abnormal AI Email Security Now Available in Japan via SoftBank

Abnormal AI, the AI-native behavior security company, today announced that SoftBank Corp. (“SoftBank”) began offering its email security platform to enterprise customers across Japan.

Abnormal has operated in Japan since 2025, led locally by Country Manager Kei Mitsuyama. Through SoftBank’s offering of Abnormal’s email security solution, Abnormal will be able to reach a broader base of enterprise customers across the country through SoftBank’s enterprise sales network.

Phishing and other social engineering attacks, including business email compromise (BEC), are growing more sophisticated across Japan. According to Japan’s National Police Agency, citing Council of Anti-Phishing Japan data, phishing reports in 2025 totaled 2,454,297¹, underscoring the scale of the challenge enterprises face.

Abnormal is designed to counter these attacks with a proprietary behavioral foundation model that builds a deep, environment-wide understanding of what’s normal across email, identities, vendors, and workflows in an organization. When something deviates from that baseline, such as an attack, a compromise, or a risk, the Abnormal Behavioral Security Platform is designed to detect and remediate it automatically, including many previously unseen threats, without relying primarily on rules or signatures.

Unlike traditional email security gateways, which typically require rerouting mail through MX record changes, Abnormal AI connects directly via API to Microsoft 365 and Google Workspace, eliminating the need for MX changes and enabling quick, smooth deployment.

As Abnormal AI’s solution is offered to SoftBank’s enterprise customers, adoption is expected to accelerate among companies already running Microsoft 365 or Google Workspace.

Supporting Quotes:

Evan Reiser, CEO and Co-founder, Abnormal AI:

“Abnormal’s mission is to protect organizations from sophisticated, evolving attacks, no matter where they operate, and that mission is exactly why we’re deepening our commitment to Japan through this partnership with SoftBank. We’re pleased to bring our behavioral AI platform to more Japanese enterprises at a time when email-based attacks keep growing more advanced. Together with SoftBank, we’re giving Japanese enterprises the protection they need to stay ahead of evolving cyber threats.”

Kei Mitsuyama, Japan Country Manager, Abnormal AI:

“Japanese enterprises tell us the same thing we hear in every market: legacy, rules-based tools often struggle to keep pace with today’s email attacks. SoftBank’s deep customer relationships and trusted position mean we can now bring that protection to organizations we couldn’t reach on our own, through the partner they already rely on for their security needs.”

Tomochika Morita, Senior Vice President, Enterprise Business Unit, SoftBank Corp.:

“Our customers’ security challenges are becoming more complex every year, and countermeasures against email-based attacks are an urgent priority. Through this partnership with Abnormal AI, we will strengthen our security portfolio and provide solutions our customers can use with confidence.”

To learn more about the sales of Abnormal AI’s email security solutions in Japan, contact your SoftBank enterprise representative or visit abnormal.ai/ja.

¹National Police Agency, citing Council of Anti-Phishing Japan data, ‘Threat Situation in Cyberspace, Reiwa 7 [2025]: https://www.npa.go.jp/publications/statistics/cybersecurity/data/R7/R07_cyber_jousei.pdf

Disclaimer

Statements in this communication that express beliefs or expectations about future events or performance are forward-looking in nature and are based on current assumptions. Actual results may differ materially. Abnormal AI undertakes no obligation to update these statements.

About Abnormal AI

Abnormal AI stops cybercrime with AI, protecting more than 4,500 organizations, including more than 25% of the Fortune 500, as of July 2026. The Abnormal Behavioral Security Platform spans email, identity, and insider threat, built on proprietary Behavioral AI models designed to detect the attacks legacy, rules-based tools cannot. Learn more at abnormal.ai.

Media gallery