Business email compromise losses have surged 70% since 2020 as attackers use generative AI to create hyper-personalized attacks at unprecedented speed. Traditional security tools fail against these threats because AI-generated phishing emails pass authentication checks, contain no malicious links or attachments, and originate from legitimate or compromised accounts. Organizations face supply chain attacks where threat actors copy existing email threads and mimic authentic communication patterns to request payment redirections or credential updates.

Defensive AI must operate autonomously to match attack velocity, analyze behavior continuously rather than relying on signatures, and integrate seamlessly into security workflows.

In this insightful session, Margaux Tawil, senior enterprise engineer at Abnormal AI, will discuss:

• Bypassing AI guardrails through prompt engineering to generate phishing content;
• Identifying zero-day attacks that evade signature-based detection;
• Automating incident response through API-integrated remediation across mailboxes.