Skip to main content

Dangerous malware is up 86%: Here’s how AI can help

Jolly Roger - glitch malware concept
Image Credit: traffic_analyzer/Getty

Join us in Atlanta on April 10th and explore the landscape of security workforce. We will explore the vision, benefits, and use cases of AI for security teams. Request an invite here.


We’ve all heard that data breaches and other cybersecurity threats are on the rise, but the latest numbers, from SiteLock’s 2022 Security Report, are truly staggering: “High-severity malware” is up 86% year-over-year. High-severity threats are those that can lead to stolen data, loss of customer trust, and damage to reputation and brand. And unfortunately, these threats often impact users when they’re doing nothing more than visiting a website or clicking on a link.

More specifically, this includes the likes of malware downloads, phishing, redirects to malicious websites and code injection. In other words, it’s becoming increasingly easy for hackers to target innocent users and wreak havoc on their lives and businesses.

As we all spend more time online in the post-COVID world, we’re risking more exposure to these kinds of threats. Cybercriminals are taking advantage of the work-from-home trend to target remote workers and their company’s data. They’re also using the pandemic as a cover to launch more phishing attacks, since people are understandably distracted and preoccupied these days.

Big Data and AI can find malware

Big data and AI can help us fight back against these threats, as patterns emerge and new risks are identified. An IEEE study reviewing AI-powered malware detection methods concludes that these techniques “provide significant advantages,” including in terms of accuracy, speed and scalability. SafeDNS, for example, uses “continuous machine learning and user behavior analysis” to achieve 98% precision in detecting malware.

VB Event

The AI Impact Tour – Atlanta

Continuing our tour, we’re headed to Atlanta for the AI Impact Tour stop on April 10th. This exclusive, invite-only event, in partnership with Microsoft, will feature discussions on how generative AI is transforming the security workforce. Space is limited, so request an invite today.
Request an invite

So how exactly does AI work in this context? Well, SafeDNS’s technology has built a large “database of malware,” and data is the key to AI. The more data you have, the more accurately you can train your AI models. AI then constantly analyzes this data to look for new patterns of behavior that could indicate a threat.

This is just one example of how AI is being used to help us stay safe online. As the research makes clear, we need all the help we can get. With high-severity malware on the rise, AI is one tool that can help us fight back.

What happens when malware is detected

Of course, detecting malware isn’t enough. If a machine learning model predicts that a website is hosting malware, what happens next? In the case of platforms like SafeDNS, the user is warned and redirected away from the malicious site. This happens in real time, so the user never sees the dangerous content.

This approach is at the user-level — but what about the root cause? In other words, what happens to the website that’s hosting the malware? According to SiteLock’s report, the answer is often “not enough,” as the findings reveal that 92% of infected websites are not blacklisted by search engines.

One issue is that, as CNN reports, if Google detects “persistent malware” on a site, it will block the website. However, many malicious sites are only active for a short period of time, so they may not be caught by Google’s algorithms. Moreover, Google “quarantines” rather than “blacklists” many of the sites it does detect, meaning they’re still accessible — just with a warning.

As a result, the damage can still be done, and users can still be exposed to malware. This is a serious problem that needs to be addressed. Considering the continued prevalence of data breaches, it’s clear that more can be done to protect users — and AI is one tool that can help.

Attackers are using AI, too

There’s a veritable arms race between cybercriminals and security experts, and AI is playing a role on both sides. For instance, a VentureBeat article explains that “AI-powered cyberthreats pose a major risk to enterprises, particularly those engaged in frequent communications susceptible to phishing.”

On the one hand, AI is being used by attackers to automate the process of launching phishing attacks and other malware. This means that they can scale up their operations and launch more attacks more quickly. On the other hand, AI is also being used by security experts to automatically detect and block these attacks. As we’ve seen, AI-powered malware detection is an effective way to protect users from exposure to dangerous content.

The side that fails to use AI effectively will lose the race. This is why it’s so important for security experts to continue to develop and refine their AI-powered tools. Cybercriminals are not going to give up easily, and we need to be prepared to fight back.

AI is only going to become more important in the realm of cybersecurity. As the SiteLock report makes clear, the need for AI-powered tools is only going to become more urgent as high-severity malware becomes more prevalent. We need to be proactive about developing and deploying these tools if we want to stay ahead of the curve.

This is not a problem that’s going to go away anytime soon, and we need to be prepared for the long haul. Cybersecurity is a critical issue, and AI is one of the keys to keeping us safe.

Valerias Bangert is a strategy and innovation consultant, founder of three media outlets, and published author.

DataDecisionMakers

Welcome to the VentureBeat community!

DataDecisionMakers is where experts, including the technical people doing data work, can share data-related insights and innovation.

If you want to read about cutting-edge ideas and up-to-date information, best practices, and the future of data and data tech, join us at DataDecisionMakers.

You might even consider contributing an article of your own!

Read More From DataDecisionMakers

VB Daily - get the latest in your inbox

Thanks for subscribing. Check out more VB newsletters here.

An error occured.