Moving People and Information Securely with Behavioral AI-Based Cloud Email Security

Tas Jalali, Head of Cybersecurity

AC Transit serves about 13 counties and nine incorporated areas in the Alameda-Contra Costa district. We serve over 158 service lines, serving about over 700 transit buses. We serve about 200,000 customers on an average weekday. There is a long-standing commitment to transit customers in the district.

When I was brought in to build this cybersecurity practice, there was an ATO attack in progress
and we badly needed to stop it because AC Transit’s a partner to a lot of public organizations. That includes pretty prestigious organizations, including Defense, Department of Homeland Security, FBI. Now when the email is going out from a reliable agency to those agencies, now we're talking about national security.

That's why preventing and stopping this attack was so critical. And we did not have time to deploy heavy, painstaking, secure email gateway. That's where we found the API-based solution of Abnormal Security. It allowed us to go and plug it in. Not only stop all these attacks, to remediate all these emails already sitting in thousands of email boxes within AC Transit.

It took us about one week to do the PoV and by end of next week, we're up and running. Not only did it help us deploy the solution pretty quick, but also helped us metastasize off the spread of the ongoing attack and stop it. We needed something that will not only use machine learning to detect these advanced attacks, but also use content and behavioral-based modeling of AI, and some recognition patterns can be used to trace advanced attacks.

We needed something more and beyond, and that's where Abnormal really came in. They had features like clawback, automatic clawback, which allowed us to prevent those ongoing attacks, which are already sitting in the email inboxes.

Abnormal has been a great partner. They have been responsive. They have been listening
to our needs and requirements. And they have made significant change based on our input. So which we really love.

I mean they have been really focused on customer service and there’s absolute trust.

Tas Jalali, Head of Cybersecurity

Ransomware, spear phishing, BECs, ATOs, they are pretty common, so you need more than simple machine learning where you are using content-based and behavioral-based technologies which can not only identify very unique communication patterns but also anomalies.

We see a lot of our executives and the board of directors is very commonly hit with phishing attacks, and then BEC attacks, on impersonation attacks, and Abnormal has been really great in detecting it. So we have pretty much 99% success with Abnormal in detecting impersonation, phishing, and BEC attacks.

So that's where we find Abnormal really doing a great job.

Tas Jalali, Head of Cybersecurity

A few other features I really like about Abnormal is that they allow us to filter promotional email. Most of the executive staff, management, they get tons of promotional emails. They're not necessarily malicious emails. So I get probably about 200-300 emails a day and are promotional emails. And if I were to filter and go through them, they’d probably be taking 20 to 30 percent of my time a day.

So with this feature, it has allowed our users to be more productive and focused on productive things