Using AI to Strengthen Cybersecurity
It’s a grim reality: Bad actors are harnessing AI to improve the effectiveness of their attacks. It’s time for IT and business leaders to put AI to work defending their data. The best place to start is to speed up attack detection.
“When it comes to discovering attacks, it’s all about the data. The faster you can analyze it, the better,” says Jackson.
“Although the sheer quantity of data that must be scanned is steadily increasing, there’s more processing power than ever — plenty for AI to discern trends and patterns that might betray a breach,” adds Healey.
AI is much on the mind of IT and business leaders who are concerned about beefing up cybersecurity — and who isn’t? In the CIO MarketPulse research survey, faster detection is the top-rated cybersecurity benefit (63%) that respondents expect to receive from AI.
50% – Faster remediation
56% – Ability to embed security software into development cycles (aka DevSecOps)
48% – Ability to improve access control by analyzing anomalies that arise in typical user activities
False positives are the bane of many a cybersecurity manager’s existence. “False-intrusion detection is a massive opportunity for recent advancements in AI, driven by almost unlimited computing power,” says Healey, noting that having many false positives can overwhelm cybersecurity staff.
“Instead of heightening their awareness of danger, a plethora of false positives can cause staff to ignore all alarms, greatly increasing vulnerability to attack,” Healey explains. In the survey, 47% of all the responding decision-makers said AI will help reduce the number of false positives. IT and security pros seem especially confident: 56% said AI will help reduce false positives.
As GenAI comes into widespread use, it will be particularly useful in penetration testing. For example, GenAI tools can be asked to create a cyberattack with a high probability of success against an organization that has a particular set of defenses.
Jackson offers another note of encouragement: “At OpenText™ Aviator Intelligence we are infusing AI into every product. This AI is invisible, because it’s embedded, so it does the work for you. And because it’s AI, the more you use it, the more you train it, so its effectiveness steadily increases over time,” she says.
Ultimately, the stakes are too high to leave AI on the sidelines, Jackson says. “You must ask yourself, ‘What’s the risk of not putting AI into a security platform?’”
