Case Study: Berkeley Group blocks advanced email attacks and saves 366 SOC hours with Abnormal AI

Protecting Sustainable Community Development with Human Behaviour AI

Berkeley Group, a UK homebuilding and community developer with 89 development sites and 2,800+ protected mailboxes, faced persistent credential phishing and business email compromise (BEC) attacks—many impersonating executives—despite Microsoft 365 and a secure email gateway. To reduce reliance on user reports and give their small SOC more time for complex cases, Berkeley Group turned to Abnormal AI, deploying its human behaviour AI via Inbound Email Security, Account Takeover Protection, and AI Security Mailbox.

Abnormal AI integrated with Microsoft in minutes and used behavioural signals and automation to block and auto-remediate sophisticated phishing that previous tools missed. The deployment took five minutes to stand up, saved 366 SOC hours in the first four months, and stopped notable attacks (18.9% of attacks stopped in the first 90 days were text-based), improving investigation speed and inbox trust while reducing manual reporting and remediation overhead.

Berkeley Group

Ash Hughes

Head of Security