As IT Business Edge notes, AI has become a crucial weapon in the cybersecurity arsenal in recent years, enabling better prediction of attacks and even helping cut down on frequency and impact of attacks.

Other benefits of AI include being able to handle vast amounts of data, to adapt quickly, to detect new cybersecurity risks early, to monitor threats in real-time and curb phishing. According to one recent survey, about four-fifths of U.S.-based organizations deem AI necessary for cybersecurity.

However, hackers enjoy the fruits of AI, too, and use the technology to pull off ever more complex cyberattacks. AI-enabled devices could also pose privacy risks. Moreover, the cost of talent in the developing AI space is high.

According to CPO Magazine, one way that AI has encroached upon cybersecurity is by hackers using “synthetic images” to extract usernames and passwords, thereby gaining access to confidential data on a device. A team of researchers has also come up with AI clever enough to fool the Captcha tests used to determine if a website is being accessed by a human.

There are other positives. As Edtech Magazine observes, machine learning can also help cybersecurity defense increase in scale as the number of endpoints increases, such as with the past year’s shift to remote learning.

Still, Matthew Carroll, CEO at cybersecurity firm Immuta, has called AI “cybersecurity’s blind spot.” Writing for GCN.com, Carroll concluded, “The long-term success of the nation’s cybersecurity policies will rest on how accurately they apply to the world of AI.”

In the meantime, IT Business Edge suggests an approach combining AI capabilities with human oversight, striking a balance to be sure the system isn’t being manipulated or relying on flawed reasoning. CPO Magazine calls for a coalition of nations, companies and cybersecurity experts to harness the potential of AI before the hackers do.