[comp.virus] Tiger Teams

XRAYSROK%SBCCVM.BITNET@VMA.CC.CMU.EDU (Steve) (09/27/89)

   Maybe I just don't understand, but I personally think the "Tiger Team"
idea put forth (by David Gursky) on this list is a little ridiculous
because:
   1) Most viruses are not spread by someone sneaking in at night and
against your wishes copying something onto your computer.  Rather,
they are usually spread voluntarily (but unknowingly) by the user
exposing the computer to foreign contaminated disks or programs.  If I
always (almost always anyway) operate within a closed system, how is
letting someone *tamper* with my computer going to help me? I'd feel
much safer just scanning for known viruses, which brings up the next
point.
   2) What corporation (or employee for that matter) is willing to
take the risk of letting someone (outsiders or corporation employees)
*tamper* with the computers which the company (and the employee)
depends upon, especially when proper operating procedures (regular
backups, etc.) will offer you very good protection?
   3) Can you guarantee that the "Team" will not do damage?  No, you
cannot.  And if they are introducing live viruses, we already know
that no one can guarantee that the viruses will be benign in every
situation (as has been discussed many times by others on this list),
or that they will not get away.
Acknowledge-To: <XRAYSROK@SBCCVM>

XRJDM%SCFVM.BITNET@VMA.CC.CMU.EDU (Joe McMahon) (09/27/89)

Dave Gursky asked about the tiger team approach. It depends on several
things:

- - Is the computer in question a computer which belongs to the installation,
  or one which belongs to the person?
- - Is the virus completely self-limiting (i.e., if the date becomes anything
  other that the date of infection, the virus removes itself?
- - Is the company willing to risk destroying this user's files and possibly
  wasting large amounts of time and money to replace them?

Apple's statement on Mac viruses is that you should never trust a
once-infected file, even if it is "cleaned up". I tend to side with
that approach.  I know that if I had been following procedures, and
some expletive-deleted from Security futzed around with my machine
behind my back, I'd be angry.  Especially if it trashed my files.

 --- Joe M.

ignatz@att.att.com (09/28/89)

In article <0002.8909261721.AA06193@ge.sei.cmu.edu> dmg@retina.mitre.org (David
 Gursky) writes:
			...
>Suppose a company has stringent rules about protecting desktop
>computers from viruses.  How do you go about ensuring the rules are
>being followed?  One thought I had was the user of "Tiger Teams".

And goes on to describe a "Tiger Team" which would prowl the halls
after-hours, looking for unsecured desktop machines which it could
then infect with an "approved" virus, preparatory to an upleasant
visit by the PC Police the next day.

Presumably, the purpose of actually infecting the machine is to
provide an object lesson to the unhappy employee careless enough to
not lock the system.  This, however, is Not A Good Idea, for many
reasons.  First, you've disrupted the productivity of a probably
useful employee for at least half a day, or more, while his/her
machine is zoned out.  Next, you're tying up one or more people
comprising the "Tiger Team"; as proposed, worse, they're having to put
in non-prime hours performing what is essentially an overhead (read
"costs money, makes none") task; you're setting up the kind of
confrontational situation that can cause stressful relations between
employees; and it's not necessary.  Not to mention that there are
other security holes that are unaddressed, such as terminals left
logged into multi-user systems which nevertheless can be used to
corrupt or destroy company data and programs.  Also, how about desktop
or cubicle multi-user and/or multi-tasking systems, such as small
Unix/Xenix boxes, VAX/VMS workstations, etc.?  Look at finding access
to these, and then corrupting them, and you'll start to see that this
is a form of sanctioned cracking which is beneficial to none, and
detrimental to all.

More useful, and actually used in many client sites I've been assigned
to, is to simply have the guard--who must make rounds anyway--also
made responsible for checking certain criteria for computer equipment.
Such things as locked access when applicable, no media left lying
about unattended, login-protected terminals (whether remote
timesharing, desktop multi-task/user, etc.) logged off whenever
unattended, etc. would be grounds for a report by the guard.  At the
same time, the unsafe condition would be corrected as well as possible
by the guard--media collected and secured, accounts either logged off
or reported to system operators for deactivation, unlocked single-user
desktop machines either locked in the office, if possible, or the
power supply secured, etc.  The same desired benefits are obtained:
the employee is made amply aware of his/her faux pas, and security is
maintained.  Anyone who's ever worked in a security environment is
aware of these and other methods; they're actually used, as I
mentioned before.

The military does make use of "Tiger Teams" that attempt to penetrate
security and leave proof of their success.  Usually, however, they are
employed in an environment where they're attempting to subvert or
circumvent active security measures, such as the deck guard on a nuke
sub that's docked, or access to a presumably secured and monitored
area.

TBC101@PSUVM.BITNET (Thomas B. Collins, Jr.) (10/01/89)

Another thought on the Tiger Teams...  It doesn't make much sense to me.
If I don't add any new software to my system at work, I'm not going to
worry about viruses.  Say I get my new system, put all the software on
it, and run a few virus scanners that turn up nothing.  I then run all
applications from my hard drive, and don't use any floppy disks.  It
wouldn't make sense for me to check my hard drive every day for viruses,
because they don't just pop up from nowhere.

If I did add software to my system, I would check it for viruses before
adding it.  I think it would make more sense for the Tiger Teams to come
in in the middle of the day, ask you to please save your work, and then
run a virus checker on your system.  If anything is found, you are
"cited" as letting a virus into your system.  If you're clean, you go
back to work, and the Tiger Team moves on.

- -------
Tom "Shark" Collins       Since ICS is comprised of 2 people, my views
tbc101@psuvm.psu.edu      are the opinion of at least 50% of the company.

okay@tafs.mitre.org (Okay S J) (10/03/89)

In VIRUS-L V2NO208 "Thomas B. Collins, Jr." <TBC101@PSUVM.BITNET> writes:
>Say I get my new system, put all the software on
>it, and run a few virus scanners that turn up nothing.  I then run all
>applications from my hard drive, and don't use any floppy disks.  It
>wouldn't make sense for me to check my hard drive every day for viruses,
>because they don't just pop up from nowhere.

You're discounting the fact that your machine could be on a network. Having
an infected machine on a network where one transfers files between machines
can be just as bad as sticking a floppy in the machine.  One shot does
not cure all

>If I did add software to my system, I would check it for viruses before
>adding it.  I think it would make more sense for the Tiger Teams to come
>in in the middle of the day, ask you to please save your work, and then
>run a virus checker on your system.

It would cause too much of a loss of productivity and interruption of
the work routine.  Night is better if you're going to do it. Plus the
public embarrasment of having ones machine checked. Seriously, its
kind of like any test for drugs or AIDS or anything like that. Its not
so much as to whether you are infected, but just the idea that it was
done. After all, why have a test done if there isn't some
suspicion...This at least would be the view of most people around
those who had their machines tested.  'Did you hear George got busted
by the Tiger Team last week?---They didn't find anything, but you
never know....'

>If anything is found, you are "cited" as letting a virus into your system.
>If you're clean, you go back to work, and the Tiger Team moves on.

What exactly does 'cited' mean? Disciplined?, public marked as a
electronic leper in the company? fired? --Now that we've established
how they would operate, what should be the penalties for those
'caught'?

Stephen Okay    Technical Aide, The MITRE Corporation
x6737        OKAY@TAFS.MITRE.ORG/m20836@mwvm.mitre.org
               'Geez...I actually have to use a disclaimer now,
                I must be getting important!'
Disclaimer:Its mine, mine, mine, mine, mine !!!!!!!!!!!!!!

V2002A@TEMPLEVM.BITNET (Andy Wing) (10/06/89)

Hi,
     I think that your average non-sophisticated user would be
offended by computer support personnel checking their personal
machine for "infection".  An alternative would be to have the
Tiger Teams simply state that they are doing "regular preventative
maintenance".  People shouldn't have problems with that.  The end
user doesn't need to know the gruesome details of a PM call.
     Actually Tiger Team duties should be assigned to a companys
regular maintenance people (with a software expert supervising
them of course).  I guess the best anti-virus protection is one
that is both transparent to the end user and in the hands of a
well trained support staff.
     The original Tiger Team idea would work best if slightly
modified.  Every football team has both an offence and a defense.
Right now the anti-viral defense really has no one to practice
against.  I think what we need is a group of developers that will
try to "bust" Gatekeeper/Flushot/etc.  These people would be
in close contact with the anti-viral developers.  The Tiger Team
would document their methods and only use benign infections.
     I guess my real concern is that anti-virus developers take
a reactive stance instead of an active one.  If I were a anti-virus
developer, I would want to encounter a new infection method under
controlled, documented conditions.  This way anti-viral SW would
be guarded against bypass methods already thought up by the Tiger
Teams.
     Also, do any anti-viral programs use the 'bad block' method
to protect themselves?  I think that idea holds some promise.

   Andy Wing     V2002A@TEMPLEVM.BITNET