News for hackers. Stuff that matters.
Re:Investigating a malware sample which sometimes fails to b... by Animats on Saturday September 24, 2016 @08:29PM (#12573254)
Malware which simply doesn't do anything for the first few weeks, or even hours, is hard to detect by running it in a VM box. Some "advanced persistent threats" are like that.
Re:Investigating a malware sample which sometimes fails to b... by meowface on Sunday September 25, 2016 @01:54AM (#12574168)
This has become really common. Most modern malware sandboxes attempt to greatly accelerate the system clock to account for this, in addition to trying to detect and intercept long sleep (or equivalent) calls.

Like most long cat-and-mouse games, modern malware families all use a combination of many different tricks and anti-sandbox/analysis techniques. The sandboxes play whack-a-mole at the same rate as malware keeps adding more moles.

Re:Investigating a malware sample which sometimes fails to b... by Tinyyy on Sunday September 25, 2016 @02:16AM (#12574226)
If the malware attempts a CPU test that maxes out the host CPU, will it be obvious that the system clock is sped up?
Re:Investigating a malware sample which sometimes fails to b... by cnvogel on Sunday September 25, 2016 @09:03AM (#12575130)
So… only speedup the clock while the software under test is sleeping (waiting on a muted, timer,…) . Stealthy software shouldnt have excessive cpu consumption, otherwise people will investigate.

It's obviously a cat & mouse game.

Re:Investigating a malware sample which sometimes fails to b... by Senji on Monday September 26, 2016 @12:06PM (#12582993)
I propose a scorched earth tactic. Refuse to run unsigned code.
Re:Investigating a malware sample which sometimes fails to b... by EGreg on Saturday September 24, 2016 @10:36PM (#12573613)
I can imagine that someday viruses can be written to have no signature to detect at all
Re:Investigating a malware sample which sometimes fails to b... by pizza234 on Saturday September 24, 2016 @11:02PM (#12573693)
This has been done (relatively) long ago.

It's called metamorphism - the entire code of the virus is continuously changed. They're incredibly complex pieces of code; in at least one case, 95% of the code was devoted to the metamorphic engine.

Note the difference with polymorphism, where only the decryption routine[s] change.

Two notable viruses have been Zmist and Simile.

Re:Investigating a malware sample which sometimes fails to b... by Animats on Saturday September 24, 2016 @10:49PM (#12573647)
Polymorphic viruses.[1] No two instances are the same. In some cases, not even the same length.

[1] https://en.wikipedia.org/wiki/Polymorphic_code

Re:Investigating a malware sample which sometimes fails to b... by Cyph0n on Sunday September 25, 2016 @12:43AM (#12574010)
Yeah, I recall polymorphic malware toolkits being sold in mid-level marketplaces back in 2010.