New security breaches in Microsoft products are revealed with
distressing regularity. Aside from dealing with the concretes, this
also raises more general issues: why is Microsoft unable (or
unwilling) to deal with these security issues, and how should a
small to medium firm that does not have the resources to devote a
part-time employee to security approach this problem? Realistically,
to what extent is a small firm actually at risk?
Let us consider the specifics first. The latest security issue is
unique in that it does not involve a virus, but an everyday feature
of Word, the use of field codes. Word uses field codes for a number
of ordinary functions, such as setting the date, or putting the name
of a document in a footer so that it is printed with the document.
The issue is as follows: someone, sends you a document to be edited.
You open the document, edit it, print it, and return it to the
sender. Unbeknownst to you, "spy" field codes in the document have
inserted documents from your hard drive or server into the document
or sent them to a web site. The original sender has "stolen" some of
your documents, and there is no way for you to be aware of this.
This exploit was first revealed on August 26. To date, Microsoft has
refused to recognize the seriousness of this problem, although
columnist Woody Leonhard reports that its PR agency has sent an
email to one journalist claiming that a "fix" is in the works (not
for Word 97, which Microsoft no longer supports, though). More ways
to use this particular field code are being published every day and
the potential damage it can do is expanding apace. For example, it
was originally thought that the sender of the document had to know
the exact name of the document he wanted to steal, but that is no
longer entirely true.
The easiest "fix" for this problem is to obtain a free utility by
Bill Coan, which you can run against any document to see whether it
contains a "spy" field. This is available at http://www.woodyswatch.com/
util/sniff or http://www.wordsite.com/HiddenFileDetector.html. If
you are already using Payne Consulting’s Metadata Assistant, this
supposedly also incorporates a fix for this problem in its latest
release.
Will This Actually Happen to You?
This is a widely published exploit that does not require any
programming skills other than a moderately sophisticated knowledge
of Word. It is not and cannot be picked up by any virus scanners
because it is not a virus.
Therefore, any deal or case in which the stakes are high enough
poses a risk that someone will try to steal sensitive documents. To
some extent the question "how likely is it that this will happen" is
irrelevant, since it only takes a single instance for you to lose a
big case, be sued for malpractice, etc. Other types of disaster are
not very "likely" either, but you still have insurance to protect
you. In this case, the "insurance" is free: get the utility and run
it against every file sent to you by anyone outside your firm.
Other Security Issues
The risks posed by Word’s track changes function have been
recognized for several years, and utilities exist to eliminate the
danger posed by metadata. This risk is quite serious and actually
happened in at least one instance I am aware of. If you open a Word
document that had tracked changes turned on using WordPerfect (or
any text editor), you see all the comments and changes. One firm
received a document written in Word, opened it with WordPerfect and
noted the following comment concerning one passage: "Jim, do you
think we can get away with this language." Needless to say, it was
trivial for the attorney who opened the document to say in the
course of negotiations, "now, you know I won’t let you get away with
that language."
Again, utilities exist to minimize this danger, and as a matter
of policy, documents should never be sent out of the firm without
accepting all tracked changes. If you were really paranoid, you
could open every Word document in WordPerfect before you send it out
into the world to make sure it is safe.
Internet Explorer
Internet Explorer occupies a special place in the pantheon of
security risks because it is so tightly integrated into Windows (can
you say "antitrust"?.....). In addition to Microsoft products, other
software programs are increasingly requiring that Internet Explorer
must be installed, even if you don’t use it (e.g., PC Law, Amicus
Attorney, Summation, and others). IE security breaches will affect
you even if you don’t use it.
Therefore it is critical to keep IE updated. Unfortunately,
Microsoft's "critical" updates are not always reliable and in some
cases can lead to re-opening old security holes. A Microsoft
knowledge base article notes that one "fix" is to tell IE not to
trust content from Microsoft! This gives you control over what you
install. To do this, in IE, click Tools | Internet Options |
Content. In the Certificates section click Publishers | Trusted
Publishers. If Microsoft is listed, click on it and click Remove. In
the future, as Microsoft implements its new license provisions that
allow it to change the configuration of your PC without letting you
know about it, this will be even more important. You may also want
to disable the auto-update "feature" in WindowsXP. To do this, go to
Control Panel | Administrative Tools | Services and change
Auto-Update to manual.
Outlook Issues
The two main ways that viruses spread at the present time are
through Internet Explorer and Outlook. Microsoft's response to these
issues has been to lock down Outlook through a draconian security
patch that serious inhibits Outlook's ability to integrate with
third-party programs such as the PalmPilot. You now have to tell
Outlook that you do want to do the link and for a maximum of 10
minutes.
If you are using Outlook with Exchange Server, there is a patch
that enables an administrator to disable this warning. If you are
using Office XP, you might want to get Ken Slovak’s utility that
lets you selectively re-enable options that Outlook outlaws en
masse. See www.http://www.slipstick.com/ files/attopt.zip
The standard methods of protecting against virus infection (in
addition to an anti-virus program that is updated very often) have
been 1) to close the viewer pane in Outlook; 2) never to open an
attachment that you are not expecting to receive.
However, with the spread of viruses through IE-related holes,
this is no longer sufficient. Even more serious, the newest viruses
spread by sending themselves to everyone on your e-mail list. Thus
you can receive a virus in what appears to be an e-mail from someone
you know.
In response to all these issues, an entire cottage industry has
grown up to remedy the security problems with Microsoft products.
Two of the best sources are Woody Leonhard's "Woody's Watch" site (www.woodyswatch.com)
and his various newsletters, and Sue Mosher's Outlook site,
Slipstick, at www.slipstick.com. These are worth checking regularly.
Why Doesn't Microsoft Fix It?
The obvious question is: why can't (or won't) Microsoft fix all
these problems? Until recently, Microsoft's main stress was on "ease
of use." Since this ease of use was implemented through the same
procedures used by virus writers, Microsoft regarded its security
holes as features or assets rather than as problems. More recently,
Bill Gates announced his goal of providing "trustworthy computing."
Aside from whether or not you can take Microsoft pronouncements as
good coin, there is a serious structural problem here. To truly
eliminate the rampant security breaches, the basic code of Windows
and other Microsoft products will have to be re-written from scratch
and will almost certainly be incompatible with all previous
versions. This is not only a massive undertaking, but likely to
engender the major problem that all previous versions of any
software you use will no longer work. In short, implementing
"trustworthy computing" impinges on Microsoft's ability to maximize
its profits, and is therefore not likely to happen.
What Is Realistic?
It is safe to say that a firm which does not require
login-passwords is unlikely to take a serious approach to protecting
its documents from intrusion on the grounds that "it's too much
work." There is a realistic core to this argument: it is too much
work for a small firm in the sense that a serious approach to
security would require devoting a at least a part-time staff member
to it. Yes, you can do this yourself on a haphazard basis, but
remember Red Adair’s adage: "if you think hiring a professional is
expensive, try using an amateur."
Rather than simply ignore the issue, firms might consider hiring
a consultant to come in on a regular basis – say, a half-day a month
– and go over all new security issues as they pertain to the firm.
This could also be an occasion to increase user awareness (there is
no substitute for on-going security and anti-virus training). In
addition, the consultant could be "on retainer" so that you get a
priority response in the event of a particularly serious new virus
attack, or the actual infection of your system. In short, take the
"retainer" approach that is similar to the way attorneys deal with
having experts or other attorneys specialized in certain areas "on
call" so that you know they will be available when needed.
Five Years of Computer News
This issue marks the fifth anniversary of Computer News for Law
Firms. Many of our articles have been syndicated via the
Technolawyer network and reprinted in publications reaching hundreds
of thousands of readers. Past articles are posted on the Heckman
Consulting web site at www. heckmanco.com. Some are outdated, but
those on general topics such as why use Case or Document Management
programs still read well.