acm-header
Sign In

Communications of the ACM

ACM Opinion

AI Needs to Prevent Cyberattacks, Not Just Detect Them


View as: Print Mobile App Share: Send by email Share on reddit Share on StumbleUpon Share on Hacker News Share on Tweeter Share on Facebook
Illustration shows an exclamation point inside a triangle, denoting an alert due to cyberattack.

AI can help automate cyberattack prevention, but not many businesses are using it in this way.

With cyberattacks becoming more common, many are hopeful that artificial intelligence (AI) can play a greater role in bolstering their defenses. The good news is that AI is increasingly helping large organizations to detect cyberattacks by learning a company's normal network activity and detecting aberrations.

However, while AI has made great inroads in detecting cyberattacks, it remains underutilized in preventing cyberattacks. And, it appears attackers are increasingly using AI, too, automating tasks to probe firewalls, searching for the right combination of attacks that will get through, and using machine learning to compose more convincing phishing emails.

From Fortune Eye on A.I.
View Full Article (May Require Paid Registration)


 

No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account