Traditional Security Practices Should Not Be Neglected with The Rise of AI

By Thomas LaRock, Head Geek™, SolarWinds

Credit: ID 116445150 © Tampatra1 | Dreamstime.com

Artificial Intelligence (AI) and machine learning (ML) are slowly being implemented into our everyday lives. This “rise of the machines” is unsurprisingly met with concerns, such as fearing AI will take over certain human-operated jobs. However, when used in collaboration with traditional processes, AI and ML will enhance roles and processes, particularly involving cybersecurity.

Today’s hyperconnected society—where almost everything is “smart” and connected to the network—comes hand-in-hand with a huge influx of data, which can cause a gap in security management. To help plug this gap, AI is used to sift through the masses of information which otherwise would be difficult to manage and vulnerable to human error.

IT professionals can leverage AI technology to filter and manage large amounts of information to detect suspicious behaviour. ML models learn from previous data entries, creating intelligence that can notice, predict, and prevent security breaches. This behind-the-scenes processing frees up time for IT professionals to manage higher priority tasks while being more innovative in the development and implementation of new technologies.

When used in a collaborative sense, AI has the potential to take away those tedious tasks from IT professionals, letting them focus on other priorities, such as business operations and network security.

Human management is still essential

The cyberthreat landscape continues to change rapidly, and some aspects of that landscape require human intervention now more than ever before, as cybersecurity breaches are growing in both frequency and severity. According to the Ponemon Institute’s annual pulse-check of global cybersecurity, the average cost of data breaches rose once more this year, with the number of “mega breaches” of more than 1 million records nearly doubling in the past 12 months. That suggests our current approach to cybersecurity still struggles to keep up with the sheer pace of change in the digital landscape.

Many of the biggest threats come from within: research has shown that more than half of businesses have experienced insider threats. Although these often come down to negligence or accidents rather than malicious attacks, it shows that properly trained employees are still an essential cog in a secure business.

Even though machines and systems can be highly effective at preventing suspicious behaviour, they are not great at training staff to adhere to agency policies or practice strong overall security hygiene. It’s important that companies rely on security managers to train employees on everything from potential attack techniques to simple daily habits that can help protect agency networks.

At the same time, however, AI can play a role in protecting employees from themselves. If a business can predict various events where carelessness may compromise security—such as bringing infected USBs to work, or sending sensitive documents to personal email accounts—IT can use those scenarios to “train” AI on what to look out for. From there, the AI systems should be able to gain greater insight over time into how employees behave, with feedback from IT managers helping them learn what particular heuristics to look out for with greater and greater accuracy.

That isn’t to say AI will replace cybersecurity pros at the front line. Human input still plays a major role in not only giving feedback to AI on what constitutes a risk and what doesn’t (helping to minimise false positives), but also making judgement calls on more complex or sophisticated threats—especially those an AI hasn’t seen before. When faced with a zero-day threat with little precedent, for example, human ingenuity and inference remain the best bet for quickly comprehending and containing things before a major breach occurs. But AI will help cybersecurity pros better understand situations and respond faster, even in these more dire scenarios.

Traditional practices are just as important now as ever before

Keeping a strong security foundation is essential and traditional solutions are just as critical as they were ten years ago. Configuration management and user device monitoring programs should remain at the core of business’ initiatives. AI’s role here is to separate the wheat from the chaff, identifying the critical signals that may indicate a threat out of the noise of day-to-day alerts and reports. That’s something software can do far more efficiently than humans—and a job few IT managers will hesitate to pass on for AI assistance.

Traditional network monitoring programs can analyse huge volumes of data, making them ripe for tie-ups with AI technologies. Using AI to analyse and process network data not only allows for more accurate automated threat intelligence alerts and contextual insights for managers to act on, but it also allows for automated responses, such as rediverting traffic or shutting down certain compromised network segments as soon as a breach becomes apparent. In situations where the difference between a security scare and a massive breach can come down to just minutes or seconds, that instantaneous response could prove vital to a company’s ongoing operations.

Although AI is now becoming a common business tool, the need for human collaboration remains essential. In fact, this partnership could potentially prove to be the silver bullet in the ongoing fight against cybersecurity threats—though, as with any such prediction, we’ll have to wait and see.   

Join the newsletter!

Or

Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Error: Please check your email address.
CSO WANTED
Have an opinion on security? Want to have your articles published on CSO? Please contact CSO Content Manager for our guidelines.

Tags SolarWindsmachine learningartificial intelligence (AI)

More about IBM

Show Comments

Featured Whitepapers

Editor's Recommendations

Solution Centres

Stories by Thomas LaRock

Latest Videos

More videos

Blog Posts