GUEST ESSAY: The role ‘deep learning’ AI can play relieving security teams of debilitating stress

By Karen Crowley

The cybersecurity profession can be very rewarding, but at the same time quite taxing.

Related: Equipping SOCs for the long haul

In fact, stress factors  have risen to where some 45 percent of the security professionals polled in Deep Instinct’s third annual Voice of SecOps report said they’ve considered leaving the industry altogether.

Ransomware is at an all-time high; attackers are as elusive as ever. Thus the job of detecting an active adversary and stopping them before they can do material damage has become extremely difficult.

Some 91 percent of respondents reported feeling stress in their security roles, of which 46 percent stated that the level of stress had increased in the past 12 months.

Productivity disruptor

A significant proportion of security pros concede that stress is negatively impacting their ability to do their daily tasks at work; this is the result of a number of variables including:

•A gap between the number of qualified candidates to fill positions and experienced staff members; skilled security personnel are often poached for higher wages and larger responsibilities.

•An overwhelming number of security alerts leading some organizations to turn off warnings altogether.

•Elusive adversaries who continually re-invent new ways to execute attacks.

•Newly discovered software vulnerabilities and misconfigurations increasingly getting exploited before the organization has a chance to fix them.

Above all, the core exposure derives from an increasing number of unknown threats, according to a Divisional Head of Cybersecurity Compliance at a global motor manufacturer:

“The number of unknowns is increasing. The criminals know their existing malware signatures can be detected, so they are constantly looking to find new ways to attack. It’s like they’ve got Harry Potter’s invisibility cloak. We can never switch off.”

Hero mentality

Senior security leaders, i.e. CSOs and CISOs, need to be able to convey the risks that their teams face, especially to board members who can easily get lost in explanations of the endless technical nuances.

And the more senior the cybersecurity role, the more stressful the job. Amongst senior security leaders, the top stress factors were:

•Securing a remote workforce.

•Digital transformation affecting security.

•The threat of ransomware.

A UK-based CISO at a large police force puts it this way:

“We are too reliant on the hero mentality – we have some people who are working 16-18 hour days at times. That’s not sustainable, and we certainly shouldn’t be expecting people to put in those kinds of shifts as a part of our capability. They’ll burn out.”

Taming complexity

Here are a few ways security leaders can work to reduce stress:

•Lower the volume of alerts and reduce false positive rates. Overworked SOC teams have difficulty focusing  on what really matters.

•Pull from resources from other departments, such as IT or even finance, to put an emphasis on securing the organization.

•Create clear goals and measurements of success; help security teams justify resource expenditures.

•Foster a culture of reward and positivity.


There is a great amount of discussion around AI for use cases in cybersecurity. Our survey found that 82 percent of respondents would rather depend on AI over humans to hunt threats, and 53 percent agreed that greater automation is necessary to improve security operations.

However, not all AI is created equal.  While machine learning has improved automation, it does not go far enough to make significant differences for SecOps teams.

By comparison, deep learning has been proven to provide a more preventative cyber posture for organizations. This can reduce alerts and false positives, and improve detection of actual threats bypassing controls today.

Overall, deep learning has been seen to improve not just the speed and scale of cybersecurity solutions, but the welfare and impact of security teams.

About the essayist: Karen Crowley is the director of product marketing at Deep Instinct, a New York City-headquartered supplier of a purpose-built, deep learning cybersecurity framework.

Share on FacebookShare on Google+Tweet about this on TwitterShare on LinkedInEmail this to someone