Deep Instinct: A New Way to Prevent Malware, With Deep Learning

[Guest]

Deep Instinct, founded by Guy Caspi and Eli David, Israeli Defense Force Cybersecurity veterans, applies artificial intelligence Deep Learning algorithms to detect structures and program functions that are indicative of malware.

Deep Instinct: A New Way to Prevent Malware, With Deep Learning : Read more

 

[alextheblue]

Yes, switch on Skynet! It will squash this malware.

 

[Osama68]

That's sounds very similar to what Cylance does. Cylance does not use a Sandbox as the article suggests.

 

[michaelzehr]

As alluded to, the results would be more impressive if they didn't include the training samples. Additionally while the article mentions low false positive, it doesn't mention a rate. It sounds like it's still based on analyzing the code, not the behavior, which means it can still be beaten. (Though the network analysis one sounds like it will be behavior-based.) But until we see more results, we don't know if it's a neural net that recognizes malware or a neural net that recognizes applications that do a lot of network activity.

 

[Digmeahole]

Yes, switch on Skynet! It will squash this malware.

LMAO!

 

[gibbousmoon100]

I'll never be impressed by a high success rate that isn't accompanied by a concrete and unconcealed low false positive rate. The necessity for both is common sense, imo. Why isn't it talked about more in this article?

 

[gibbousmoon100]

Accidentally upvoted my own comment, and now it won't let me downvote it. But why are we allowed to upvote our own comments to begin with?

 

[Urzu1000]

Like others have stated, they really need to include false positive statistics to make this worthwhile.

 

[HSelden]

They did include some very low false positive rates, but I couldn't publish them. At least in a uniform test, it beat all other competitors - if you blow up the image you can see the mobile test "false positives. "

I agree, the proof is real world testing, and what would Deep Instinct do with with evolving malware. Could be very interesting - time will tell. Is there an inherent structure or activity to malware that is part of it being malware, so cannot be readily disguised? Did Deep Instinct find this? If malware didn't have these behaviors or elements, it wouldn't be (malicious) malware? How could this be detected?