Darktrace, a global leader in cyber security AI, today announced that its Antigena Email product has added an early warning system, allowing members of the Darktrace community to contribute and benefit from insights gleaned from across the fleet. This new capability is now available to Antigena Email users and includes the extension of anonymized, learned domain behavioral profiles across Darktrace’s expansive and diverse group of global customers.

“Darktrace stops all kinds of cyber-attacks against organizations in every sector in over 110 countries globally. That represents a huge bank of knowledge about how malicious payloads behave in the very earliest stage of a cyber-attack,” commented Jack Stockdale OBE, Darktrace CTO. “Antigena Email has now realized the vision of leveraging collaborative, anonymized insights to leave attackers with nowhere to hide.”

Ninety-four percent of cyber-attacks begin in the inbox. As organizations continue to rely on email as a primary workplace collaboration tool and attacks become increasingly novel and sophisticated, email security technologies that rely on behavior rather than threat intelligence become more imperative.

Darktrace’s Self-Learning AI observes emails to build bespoke behavioral profiles for each customer and leverages these behavioral profiles, rather than a ledger of binary ‘good’ or ‘bad,’ to accurately determine whether each email belongs in a recipient’s inbox. Antigena Email uniquely analyzes domains within email addresses and links in email bodies and attachments to evaluate their popularity and typical presence in the inbox.

Now, when Antigena detects unusual domain behavior in a customer environment, a supplementary interpretation can be made by comparison with this new fleet-wide version of the behavioral profiles. This new functionality can lead to increased suspicion, for example, of a potential account compromise when a fleet-wide popular domain suddenly strays from its usual behavioral patterns – even in a trusted supplier or vendor.

This update recently allowed Darktrace to stop a phishing campaign sent from a compromised government account in South America that was soliciting fake philanthropic donations. Although the government domain was legitimate, the attacker had inserted their own “reply-to” address into the email headers. This address had zero domain precedent locally or globally and, in combination with other indicators, led Antigena Email to flag this email as suspicious.