I have data breach fatigue. Every day, my inbox is flooded with dozens of emails about the newest data breaches and what causes them. Five years ago, I took note of every company listed and the mistakes made that led to the breach. Today, I barely skim them. How many times can I read that a hotel chain, restaurant chain, retail chain, etc. were victims of a data breach without my eyes glazing over?
It’s not that data breaches and cyber incidents aren’t important, but that they happen so often that I’ve become immune to them, even when they have a direct impact on me.
If I, a person who writes about this, am suffering from data breach fatigue, then how bad is it for those on the front line?
The CISO’s Perspective on Data Breaches
Data breach fatigue is a trickle-down issue. Chief information security officers (CISOs) feel overwhelmed by constantly putting out fires and dealing with the aftermath of a breach. The security team feels burned out. And consumers have tuned out because there seem to be so many high-profile data breaches and their lack of understanding of cybersecurity overall.
The only way to completely get rid of breach fatigue is to stop data breaches. That, of course, is not going to happen. So, the next step is to address potential incidents before they can become full-blown threats. However, today’s cybersecurity landscape already leads to overworking and burnout. Asking people to handle more is not going to solve the problem.
Enter artificial intelligence (AI) and machine learning (ML) as a potential solution.
“You can control your own company’s security posture,” Ani Chaudhuri, CEO with Dasera says in an email interview. “Invest in solutions that automate. Also, invest in solutions that federate. Investing in solutions that improve workflow across other teams extends the security team’s bandwidth.”
The Perils of Too Many Data Breaches
No one should lose sleep because of their job, but according to a study from Forrester, a third of cybersecurity professionals are doing just that. And nearly all of them feel like it’s their fault when a breach does happen within their organization. This heavy load of responsibility is chasing too many valuable and skilled employees out of the industry. Without an experienced staff, the threat of cyber incidents increases and the cycle begins again.
On the consumer front, we’ve become blasé to data breaches, too. Many people assume that attackers have already stolen and sold their personal information so many times that they don’t see the purpose of using the most basic security practices. A Carnegie Mellon study found that only a third of consumers change their password when alerted about a breach — eventually. It can take users months to get around to doing it. And that’s if they even know they might be impacted. Instead, they shrug off alerts.
AI and Cybersecurity
First and foremost, it must be made clear: AI is not a magic solution. The technology will not solve all of your organization’s problems, but it should give CISOs and their teams the boost needed to break the fatigue cycle.
We can ‘teach’ AI to do the most mundane (and mind-numbing) tasks, some of which are involved in catching data breaches before they start. Using machine learning algorithms, AI can quickly read logs and track the network for anomalies. Too often we downplay the importance of checking logs, but this is where your team will find unusual activities occurring within your system. It’s one of the first points of attack. But it is time-consuming and takes skilled people away from other urgent jobs. Many organizations look at AI as a way to address the skills gap, and that’s true at the entry-level positions. AI cuts down the number of false positive alerts, freeing up time.
Credential stuffing has become a favorite tool of threat actors. They steal an insider’s credentials and open the doors to the entire network without easy detection. Data breaches can start from there. Multiple clusters within the network can be breached long before the security team knows about it. AI can recognize the habits of employees and use behavior models to detect something out of the norm, such as a regular nine to five employee who is suddenly accessing a sensitive database at 3 a.m. or a California-based employee who appears to be logging in from Ukraine. It’s all about pattern recognition and the ability to sniff out something that seems off.
AI Isn’t Perfect
While AI can take on many tasks of the security team and offer a more targeted approach to mitigation and defense strategies, there is a downfall that could keep CISOs up at night. AI introduces another attack vector for cyber criminals. Attackers can use AI as a disinformation tool and turn data into a weapon. An attack that takes advantage of a vulnerability in AI can lead to data breaches that are far more difficult to detect. The tool that was supposed to decrease data breach fatigue could actually end up making that fatigue worse.
AI security isn’t going to replace CISOs and security teams in the war against data breaches. Instead, it will work with humans as a way to address potential threats at the earliest stages. At best, it makes detection, protection and mitigation easier and less stressful.