November 7, 2005
Link By Link
Beware Your Trail of Digital Fingerprints
By TOM ZELLER Jr.
http://www.nytimes.com/2005/11/07/business/07link.html?pagewanted=print

IT hardly ranks in the annals of "gotcha!" but right-wing blogs were buzzing
for at least a few days last week when an unsigned Microsoft Word document
was circulated by the Democratic National Committee. The memo referred to
the "anti-civil rights and anti-immigrant rulings" of Samuel A. Alito Jr., a
federal appeals court judge who has been nominated to the Supreme Court by
President Bush.

The stern criticisms of Judge Alito rubbed some commentators the wrong way
(Chris Matthews of MSNBC called it "disgusting" last Monday). But whatever
the memo's rhetorical pitch, right-leaning bloggers revealed that it
contained a much more universal, if unintended, message: It pays to mind
your metadata.

Technically, metadata is sort of the DNA of documents created with modern
word-processing software. By default, it is automatically saved into the
deep structure of a file, hidden from view, with information that can hint
at authorship, times and dates of revisions (along with names of editors)
and other tidbits that, while perhaps useful to those creating the document,
might be better left unseen by the wider world.

(If you use Microsoft Word, open a document, go to the File menu and choose
Properties. You should see some metadata. Third-party programs are available
that will crack open even more.)

According to some technologists, including Dennis M. Kennedy, a lawyer and
consultant based in St. Louis, (denniskennedy.com), metadata might include
other bits of information like notes and questions rendered as "comments"
within a document ("need to be more specific here," for example, or in the
case of my editors, "eh??"), or the deletions and insertions logged by such
features as "track changes" in Microsoft Word.

"If you take the time to educate yourself a little and know the issues," Mr.
Kennedy said, "you can avoid problems pretty easily."

With the Alito memo - which was distributed on a not-for-attribution basis,
with no authors named - the D.N.C. was a little sloppy.

Mike Krempasky, a conservative blogger at RedState.org, mined the document's
metadata and came up with juicy, code-cryptic tidbits like this (bold added
for emphasis):

{lcub}o:Author>prendergastc{lcub}/o:Author{rcub}

Or this:

{lcub}o:Company>DNC{lcub}/o:Company{rcub}

"The technical wizards at the Democratic National Committee never got the
'don't forward Word documents' memo," Mr. Krempasky wrote, eventually
identifying "prendergastc" as Chris Prendergast and "adlerd," which also
showed up in the metadata, as Devorah Adler - both members of the D.N.C.

The metadata also coughed up a file creation date of July 7, 2005, which the
detectives at RedState.org identified as being "just after O'Connor
resigned."

None of these amounted to earth-shattering revelations, of course, but taken
together they offered a level of detail into the Alito memo that the D.N.C.
had not intended.

Josh Earnest, a spokesman for the Democratic committee, pointed out that the
origins of the document were never really a secret, even if it was
circulated as background material that was not intended to be sourced.

"Based on the fact that the D.N.C. was known to be circulating the
document," Mr. Earnest said, "I'm not sure that RedState is breaking any
news here."

Still, metadata and other document gaffes have tripped up other
organizations, sometimes with more embarrassing results.

Just two weeks before the Alito memo, the United Nations issued a
long-awaited report on Syria's suspected involvement in the assassination of
Lebanon's former prime minister, Rafik Hariri. It was a damning report for
Syria by any standard, but recipients of a version of the report that went
out on Oct. 20 were able to track the editing changes, which included the
deletion of names of officials allegedly involved in the plot, including the
Syrian president's brother and brother-in-law.

A similar gaffe embarrassed the network software company SCO Group in 2004,
when it filed suit against DaimlerChrysler for violations of their software
agreement. A carelessly distributed Microsoft Word version of the suit
revealed, among other things, that the company had spent a good deal of time
aiming the suit at Bank of America instead. "It just sort of made it look
like they were looking for the easiest target," Mr. Kennedy said.

At about the same time, California's attorney general, Bill Lockyer, floated
a letter calling peer-to-peer file-sharing software - long the bane of the
entertainment industry's interests - "a dangerous product." But a peek at
the document's properties revealed that someone dubbed "stevensonv" had a
hand in its creation.

Vans Stevenson, a senior vice president with the Motion Picture Association
of America, said later that he had offered input on the document but had not
written it.

"California AG Plays Sock Puppet to the MPAA," was one blogger's response.

The issue increasingly nags at the legal system, as lawyers become aware of
the advantages of requesting discovery of the metadata buried in
word-processed documents (or debate the ethics of scrubbing the metadata
from a file before turning it over to the other side).

"If I get a piece of paper, all I see is a piece of paper," Mr. Kennedy
said. "With an electronic document, there's potentially a lot more there."
He noted that at a recent conference on electronic discovery, an Oregon
lawyer complained that judges there tended to rebuff requests for the
electronic versions of printed documents, saying the printed versions are
enough.

But for most other instances - and certainly for cases like the Alito memo -
the solutions are simple. Sort of.

Saving a copy of a document in "rich text format" (RTF), or as a simple text
file first (options in the Save menu), and then converting it into the
common "portable document format" (PDF) before circulating it is a good
tack, Mr. Kennedy said. Still, some debate remains as to whether traces of
metadata from word-processing programs like Microsoft Word are carried
through to the PDF file.

For those who want to be extra safe, several third-party tools will scrub
metadata and other information from documents, although with each new
advance in software design, the number of potential pitfalls grows.

"It only gets more complicated," Mr. Kennedy said, making sure to point out
that all kinds of documents - from spreadsheets to PowerPoint files -
contain oodles of metadata. "It seems every time I turn around I run into
something new."

Odds are that Derrick A. Max, the head of two business groups that favored
President Bush's plans to privatize Social Security, wishes he could say the
same.

On request last spring from the Senate Democratic Policy Committee, he
e-mailed testimony on the topic. The unscrubbed Word document apparently
included editing and advice from an associate commissioner of the Social
Security Administration.

"The real scandal here," Mr. Max told The Los Angeles Times after Democrats
expressed outrage over the White House's fingerprints on the testimony, "is
that after 15 years of using Microsoft Word, I don't know how to turn off
'track changes.' "



You are a subscribed member of the infowarrior list. Visit 
www.infowarrior.org for list information or to unsubscribe. This message 
may be redistributed freely in its entirety. Any and all copyrights 
appearing in list messages are maintained by their respective owners.

Reply via email to