Cue the launch of a new AI-powered threat detection product...
Should I infect this PC, wonders malware. Let me ask my neural net...
Here's perhaps a novel use of a neural network: proof-of-concept malware that uses AI to decide whether or not to attack a victim. DeepLocker was developed by IBM eggheads, and is due to be presented at the Black Hat USA hacking conference in Las Vegas on Thursday. It uses a convolutional neural network to stay inert until the …
COMMENTS
-
-
Friday 10th August 2018 02:24 GMT Wzrd1
Cue the launch of a new AI-powered threat detection product...
Yeah, I was thinking more along the lines of a new neural net guided counterattack system.
Had something similar, years ago, which had a slightly more disproportionate response pattern. Hit a threshold, start hammering back on a sliding scale beginning at 1/3 over the level of the attack. Escalating by orders of B channels, to give a hint of the era that that system was in place.
A neural net, with the appropriate model (I have a specific animal kingdom level in mind), wouldn't be that difficult, given VM capabilities available today.
-
-
Thursday 9th August 2018 10:37 GMT Allan George Dyer
A whole new world of unintended consequencies
"When it spotted the right person in front of the PC, it provided the key needed to unlock the payload so it could be executed, and hold the system's documents to ransom."
While it will make it difficult for analysts to work out the intended target, given the state of AI image recognition, the developers won't know the effective targets, either. Sure, they trained their AI to recognise Bill Gates [other multi-billionaires are available] and empty his bank account, but maybe it triggers on a dog or guacamole.
On the other hand, if you can recognise a neural network being used to provide a decryption key for an encrypted payload, you can mark it as "highly suspicious" without bothering about when it will trigger, or what the payload is.
-
Thursday 9th August 2018 17:20 GMT Claptrap314
Crypto theory disagrees
Sprinkle magic AI dust, and everything is *NeW*, huh?
How is this AI network functioning as anything other than a hash?
Certainly, such a network is more complicated to reverse than a set of openly-tested conditions. But crypto is crypto, input is input, and a hashing function is a hashing function.
What am I missing?
-
-
Thursday 9th August 2018 19:18 GMT J. Cook
Re: Double-Edged Sword of Progress
@Jay Lenovo: I see what you did there with your username. :)
Maybe we'll get lucky and it'll be a benign AI with a cuddly avatar, like PDCL or something.
But probably not. It'll probably be Skynet.
To quote a famous dinosaur movie: "Your scientists were so preoccupied with whether they could, they didn't stop to think if they should"
-
-
-
Friday 10th August 2018 02:31 GMT Wzrd1
Re: sleeper malware 4276
And given that I am from a Sicilian-American family, said routine would never be triggered.
Even after death, when fighting relatives descend to claim "inheritance".
Our children know better, I store my wealth in a specific mineral form of wealth - cobalt-60. Inherit at your own risk, as shielding gets sold off to cover estate taxes.
-