I'm impressed with their API architecture. One of the main reasons is its invisibility to threat actors trying to launch attacks. Unlike our traditional email security tools in the SEG, which attackers can easily detect before they even start emailing us, the API remains hidden until they've already begun their attack. This gives us valuable early visibility via the API, allowing us to easily pipe that data to other tools and stop advanced attacks more effectively. The improved visibility into our email infrastructure also benefits our IT teams. Using the API integration, they can now remediate issues in minutes, whereas before it could take hours. Previously, identifying an inbound cyber attack meant bouncing between several tools: one to identify the attack, another to track affected emails, and yet another to quarantine them. Abnormal's APIs streamline this process. With a single search, an IT technician can identify users who received the emails, track who clicked on them, see where the emails are located, and even delete them from everyone's inbox directly. This has drastically reduced our investigation and response time for phishing and BEC attacks, from hours to mere minutes. Compared to many other vendors we considered, Abnormal Security stands out in its ability to detect the full spectrum of email threats. While our existing Secure Email Gateway handles traditional threats like spam and malware quite well, it often misses more sophisticated attacks. The SEG relies on static indicators like email flags, suspicious file hashes, or mass recipient lists. We can easily identify and filter out emails matching these criteria, but they do little to stop targeted attacks. Here's where Abnormal Security shines. Their anomaly detection engine excels at recognizing one-off attacks, including those where a threat actor infiltrates a vendor's mailbox and manipulates payment instructions or redirects transactions. Abnormal identifies these anomalies using behavioral analysis, effectively catching threats that traditional static methods typically miss. The two main benefits Abnormal Security offers us are its ease of use and its powerful search capabilities. These features empower our internal teams to get more involved in the response process, helping us track down threats efficiently. Additionally, Abnormal's ability to stop advanced attacks significantly reduces our security team's workload. Security teams are consistently stretched thin, so minimizing wasted effort chasing false alarms is crucial. By keeping harmful emails out of user inboxes, Abnormal allows us to focus on other priorities. In summary, our primary gains from Abnormal are its effectiveness in blocking attacks and its ability to empower our internal teams, ultimately strengthening our overall security posture. Abnormal Security's AI and machine learning capabilities significantly expand the range of email attacks they can block. This is crucial to optimizing their product's performance for us. Specifically, their ability to leverage AI indicators and extensive email telemetry is critical for stopping advanced threats, like compromised mailboxes sending disguised emails. Traditional methods often fall short in such scenarios. Our primary concern is identifying emails sent by a threat actor posing as a legitimate mailbox owner. AI-powered anomaly detection proves virtually indispensable in discerning the true sender's identity. Abnormal Security has identified and prevented several such sophisticated attacks in our own experience. One remarkable example involved a vendor's seemingly legitimate email flagged as suspicious by Abnormal. Initially dismissed as a false positive by our first responders, a deeper analysis of the email's telemetry revealed subtle anomalies. The email's sudden shift to a professional tone, unlike the typically casual communication with this vendor, was one such anomaly. As it turned out, Abnormal's suspicions were correct – the vendor's account had been compromised. This instance highlights the unparalleled effectiveness of AI in detecting sophisticated email threats. By focusing on abnormalities in email behavior, AI can uncover hidden dangers that might otherwise elude traditional security measures. The deployment of AI has significantly reduced the number of internal attacks we encounter, and it has even extended its benefits beyond our perimeter. We've proactively alerted several customers and vendors about potential compromises before they even realized their systems were under attack. This proactive approach has been well-received, with many recipients expressing their appreciation for our timely intervention. Within our organization, AI has dramatically streamlined our security operations by automating the analysis of sophisticated attacks, freeing up valuable time and resources for our security teams. Abnormal Security has dramatically reduced the time our team spends resolving email incidents. What used to consume hours or even days, depending on the attack and response complexity is now handled within minutes, often by less experienced team members. This has significantly improved our efficiency and freed up valuable time for other security tasks. Although no product can eliminate attacks, we've been pleasantly surprised by the effectiveness of Abnormal Security. Initially, when we approached them with our use case and problem, we'd have been happy with a much lower catch rate. Stopping even a significant number of attacks would have been a success. But the actual results have been incredibly impressive. While some attacks still slip through, the features in Abnormal allow us to feed those cases back into their system. This feedback fuels the AI's learning process, helping it avoid repeating the same mistakes. Interestingly, the attacks that remain undetected are often difficult to define even for human analysts. They involve subtle cues that would be challenging for any AI to spot in the specific contexts we've encountered. One example involved a new customer with whom we had exchanged only a handful of emails. While this customer's account became compromised, the attacker wasn't the usual contact person. Since the AI had only profiled the communication style of the usual contact, the malicious email appeared normal compared to that limited baseline. In such cases, where the AI lacks sufficient data, even exceptional systems can be caught off guard. While no product is perfect, we're highly impressed by Abnormal's speed and efficiency in catching attacks. They've dramatically reduced the workload on our help desk compared to the past, with the results being clear and measurable. Compared to our old solutions, Abnormal Security's incident response is like night and day. With our previous SEG, identifying and remediating a suspicious email was a cumbersome process. We'd flag the email, then jump through hoops to figure out who received it and if anyone clicked on it. With different modules and separate views, it was a mess. Once we confirmed the threat, another system hunt began, pulling emails from user inboxes. It was slow, fragmented, and frustrating. Abnormal is a breath of fresh air. If we spot a threat alert on the dashboard, we simply click on it to see all recipients, where the email sits, and who interacted with it. And then, the holy grail – a single button. Click 'Remediate', and those emails vanish from user inboxes, instantly neutralized. Just a button click from issue detection to resolution in seconds. All from one screen. That's the transformative power of Abnormal Security. Something our old solutions couldn't dream of.
