Discussion Locked

Will advanced viruses elude detection?

0 helpful
14 replies
Closed to new repliesThis discussion is closed. It stays readable, but it takes no new replies.
Tell us what you think about advanced viruses eluding detection from say antivirus software, as featured in the latest Internet Security Focus newsletter. Do you agree with the author of the newsletter that viruses will advance to the point that antivirus software will no longer be able to detect them? If so, what do you think the future repercussions will be? If you disagree with the author's prediction, what are your reasons?

14 replies

by elmonk

I share the feeling of helplessness and have always been doubtful about "hunting signatures" being the right strategy for detecting attacks. Until a real breakthrough is achieved by the cited approaches my guess is that we should shift our attentionto the entry points in our networks: what do we allow? how do we determine the authentity of the originator of some code? whom shall we trust? etc.
Of course going on with regularly updated "classical" virus detection remains a good idea I'm eagerly awaiting in-depth remedies.

by harleycat63

I have put Zone Alarm with Visual Zone, Norton Firewall/Antivirus on my system. These only do so much.
Recently I found a program that actually KILLS the links automatically once a Trojan or worm is suspected (without terminating session). It is called ProPort and I highly recommend it. Once it logs the port of entry, I then add that port to Norton's port list for future attempts.

by harleycat63

I have put Zone Alarm with Visual Zone, Norton Firewall/Antivirus on my system. These only do so much.
Recently I found a program that actually KILLS the links automatically once a Trojan or worm is suspected (without terminating session). It is called ProPort and I highly recommend it. Once it logs the port of entry, I then add that port to Norton's port list for future attempts.

by laperuta

Even if the article is, obviosly, correct, I have to disagree, due to some missing details.
First of all antivirus software not only check for recognized virus pattern, but it also prevent dangerous actions like modification of boot sectors, directaccess to disks and so on.
Second, in the case of polymoric or encrypted viruses, we have to keep in mind that a little portion of code can't be nor modified nor encrypted cause it have to be executed (the loader) by the O.S. and it must be in a "clear" form, so it can be used as a pattern.
I think the fight will become harder and harder, but it never reach the extreme point where viruses can't be detected.
Don't forget that while viruses evolve, antivirus software and technics do the same.
Bye.

by rlmieth

Many times those of us who have automatic protection within our network both at the server and the workstation, still become victim to virus or worm attacks. The most difficult aspect for the administrator is making sure all the entry points are covered and not simply bridged. I recently have had to recover from a nasty little problem which has all the signs of a virus attack yet no detection occured. One of my technicians disabled the heuristics function because it slowed down the users machines. So now i am unsure if this was a virus or a coincidence which occured at 2 sites and 5 different users.

by manabsac

We, users, are in a lose-lose situation, You PAY for an A/V S/W but can lose all data because you are late by a day to update the #@** virus defn.
What is the dismal picture? A crank can screw us all (a large no. atleast) and hell all the great menof S/W industry will post an answer to that one day later. HIGHLY SATISFACTORY SITUATION (Daily get kicked in the ... and take painkiller after that)

Merlino is totally wrong that A/V S/W can detect unauthorised writes to disk. I want to ask one simple question, How many A/V S/W scan for a second virus and the third .. after it has found one in a file? None (of the major brands - not atleast McAfee)

Software Industry is the only one without ANY liability (remember Y2K prob?) Unless we allthrow the current populist O/S and its Cotorie of supporters and puppets and introduce ACCOUNTABLITY in the industry, there is no future

by donclay

This article is right on target. Operating systems need to be written to be airtight in their design with respect to security as do firewalls. But it will all come down to individual users who will need to know from who they are accepting files.

by bill harrison

I have to agree that these intrusions will become more frequent and more difficult to resist. The part that I find bothersome is the fact that when a "vulnerability" is exploited, the OS is blamed for not being "Bullet proof". I equate virus/worm authors to stone throwing punks. If one of these punks threw a rock through your front window, you wouldn't blame the glass manufacturer for that "vulnerability". I'd be surprised if you responded by covering the remaining windows with plywood. I thinka more effective "Caller ID" or "*69" feature might slow things down a little. The next stop might be public execution of these digital terrorists...

by itprofor20

The analagy to an OS being like a "glass window" breaks down easily, IMHO. A better analogy would be the front door or some secure entrance, but even that is a stretch. If people were routinely able to gain access to my house even though I locked the front door, I would definetly blame the poor locks AND take action to improve them. With OSes, I can't (generally) take any action on my own to improve the security.

I started working on mainframe computers, then UNIX, then, kicking and screaming, into PCs. Some very good PC OSes have died due to the onslaught and strategies of Micro$oft. I never had problems with viruses until M$ OSes became so prevalent.

I believe there are two basic reasons for this:

1) M$ OSes are so widely used - the deviants who write viruses are also lazy and won't waste time porting them (even if they could!) to multiple OSes.

2) M$ OSes (and applications) are poorly written from a security standpoint - all viruses attack security loopholes in M$ OSes and products. M$ quality control doesn't seem able to discover these loopholes and reacts only after a hacker or virus writer exposes them. I never ran into a virus on a UNIX system, but I am hearing that they are popping up now. The biggest advantage of an OS like UNIX is the extensive security available.

by bill harrison

Windows whether in a house or on a pc provide certain benefits.[i.e. a view or easy, widespread operation] but those benefits come with a risk. So, you improve the lock on your front door and I smash out the window and enter your house. Are you still going purchase yet another upgrade for your lock? As for the increase in virus attacks, I agree with your first point. Thanks to Microsoft and the Windows Platform, computers are now in nearly every home and millions of people have access to them. Compared to the select few [by comparison] 10 years ago. Yes, Windows 95 put "computing" into the reach of those that would never have had the opportunity otherwise. This popularity also made the market affordable, adding fuel to the fire.

by flarich

This article highlights why a new class of products - behavior-based security software, sometimes called 'Intrusion Prevention SW', is rising. These products don't try to identify malicious code via signature analysis, they sit in the kernel and analyze the behavior of code, i.e., what it is trying to do, to identify patterns of malicious behavior. Malicious code is then blocked from execution. An example of the power of this approach is seen in blocking code that tries to propagate via spamming itself to everyone in your Outlook address book (and this covers quite a few of the recent ones). Once I have deemed that this ?behavior? is unacceptable, and block it, all future permutations of these types of attacks will be prevented?without signature updates. Note that this approach is also impervious to encryption or polymorphism. As was once said, ?malware is 1% inspiration and 99% imitation?. Behavior-based tools promise to be the best defense in this environment.

by puckducker

I am curious to see if others think as I do that along with hardening the entry points of a network, the only things that will tuly stem the rising tide of virus infestation is mandated adhereance to file transfer policies, user training and wide-spread adoption of PKI for authentication.

Don't open messages from people whos identity you can't verify. It seems to me that PKI and secure authentication could go a long way toward stemming the tide.

by drw

its USAMA BIN LADEN i know it's him and his alkita buddys are all in caves writing code and taking over the internet.

by dfw666

we have been promised for over 10 years that advanced heuristics was being built into most virus scanners and even if a new virus that wasn't in the signature list it's activity would flag it for suspicion.
Well the VBScript viruses that did the rounds proved that was a bit of a myth. However the heuristics was supposed to work if turned on but the performance impact meant that people could not turn it on.

If we had less processor intensive heuristics scanning newer virsues would be recognised. One is tempted to think that broader scanning techniques like this may not have been in the best interest of the anti-virus software vendors.

Maybe that will change with new threats.

Regards,
DFW666

This discussion is closed to new replies.