The proliferation of email phishing attacks on businesses has increased exponentially given the advanced techniques and technologies that attackers have adopted to infiltrate inboxes. These attacks, such as polymorphic emails, business email compromise and credential theft using visual similarity techniques, are so time consuming that SOC and security teams often lack the resources and manpower to detect, investigate and respond to such threats in the time they demand.

As we've written about before, it takes less than 82 seconds for a person to interact with a phishing email once it penetrates the inbox. Yet, a recent survey found that 70% of organizations take more than five minutes to remove phishing emails from corporate mailboxes an alarming percentage that is more than three times longer than the average time it takes a person to engage.

Additionally, COVID-19 has emerged as an added stressor to already overburdened SOC and security teams. Now that all non-essential companies have shifted to a 100% remote workforce, the virtual perimeter that was established in the office to protect users only partially exists. Coupled with the fear of the unknown surrounding coronavirus, attackers are taking advantage by impersonating government entities, hospital systems, insurance companies and other organizations of authority. As a result, some reports estimate that phishing attacks have gone up by over 600% since the start of the epidemic.

Prior to COVID-19, organizations of all sizes and across all industries were already overrun by phishing attacks of greater magnitude and sophistication, and security teams struggled mightily to deal with the influx in real-time. Now with the addition of a scattered workforce and broken-down perimeter, phishing poses an even greater challenge. Thus, the sense of urgency surrounding the burden of email security is greater than ever.

The Evolution of Self-Learning Email Security

Self-learning email security that combines human intelligence with machine learning has always been our philosophy at IRONSCALES. In August 2018, we launched Themis, email security's first virtual SOC analyst.

Her mission was simple: to help determine verdicts on suspicious email incidents in real-time.

Since then, Themis has helped render verdicts against more than 100 million trending and zero-day phishing attacks, including those that evade traditional SEGs and scripts like polymorphism, ransomware and BEC. To do so, she continuously learns from:

• Our proprietary data labeling technology that assess millions of technical data points
• Crowd-sourced decisions made by thousands of security analysts using the IRONSCALES platform

When combined, we have created an ever-evolving ecosystem - a feedback loop - that has supplied Themis with the unprecedented human and machine intelligence needed to render highly accurate verdicts on email phishing incidents.

Today we are pleased to announce that Themis has become more than 93% accurate in detecting and resolving suspicious email threats.

As such, SOC and security teams now have the ability to have Themis classify and remove phishing threats without their involvement, or any human involvement, whatsoever. Because Themis is now fully autonomous, SOC and security analysts can measurably increase the speed and accuracy of their email security - a major milestone in email phishing risk mitigation.

The Relief of an Autonomous Anti-Phishing Analyst

As for the other 7%, because Themis is built on top of our decentralized platform, crowd-sourced human intelligence is constantly being fed into the engine, flagging and removing threats in real-time. Themis also learns from these verdicts made by the more than 1,000 security professionals using the platform, and she will continue to learn from end users who persist in reporting suspicious messages.

Overall, Themis now allows for:

• Support where needed We live in a connected global economy that works across time zones. Ultimately, this can be an added stress to security teams. Themis adds support by being the virtual email security assistant that never takes a break, sleeps or goes on vacation.
• Consistency to email security mitigation With an accuracy rate of 93%, SOC and security teams of all sizes can be confident that Themis will mitigate email security risks with consistency while leveraging thousands of analysts inputs.
• Relief to overburdened teams By not requiring human intervention, Themis can reduce risk by resolving unknown phishing attacks faster, freeing up time for security teams to focus their efforts on other items

Ultimately, the wide-range of email security risks that SOC and security teams manage on a daily basis has increased exponentially, especially in light of our new normal remote work, personal devices, the cloud, unprotected networks, and of course, the diminishing returns of legacy email security tools and protocols. We’re working across time zones, with multiple distractions, all while trying to maintain a sense of normalcy. But even if we weren’t in the COVID-19 new normal, SOC analysts would still be facing unprecedented email threats at rates they can’t keep up with.

With Themis, security teams can harness the power of a self-learning autonomous SOC analyst on top of our decentralized, crowd-sourced engine to tackle advanced phishing attacks and relieve the pressure on security teams with a 93% accuracy rate. In fact, Themis taught herself about dozens of COVID-19 phishing scams autonomously in the past few days, just seconds after they appeared.

But our work won’t stop here. IRONSCALES will continue to strive for a higher accuracy rate with our autonomous SOC. And even though artificial intelligence may never reach 100%, our goal is to have Themis become as accurate as possible so that SOC and security teams can have the utmost confidence and relief in their fight to contain advanced phishing threats.

Don't wait to see how Themis can be your team's email security partner. Harness the Power of Now by learning more about here.