• Home /
• News /
• 2003 /
• May /
• News in Brief; Use of Data Collection Systems Is Up Sharply Following 9/11

News in Brief; Use of Data Collection Systems Is Up Sharply Following 9/11

01 May 2003

In the 20 months since Sept. 11, 2001, little-known government and commercial databases that track the movements and
backgrounds of everyday Americans have steadily ballooned.

Developed as counterterrorism tools, the systems are aimed at bridging gaps in information that let the 9/11 hijackers slip past
law enforcement. But they also make it easier for the government to gather information about American citizens who aren't
suspected of anything criminal.

Public attention has focused almost entirely on two "data-mining" projects that have drawn objections from privacy advocates: a
new airline-passenger profiling system known as Capps, intended to block suspected terrorists from flying, and the Pentagon's
Total Information Awareness program, aimed at detecting patterns of terrorist activity. Concerns about Total Information
Awareness flared again this week, when the Pentagon said it had renamed the initiative Terrorist Information Awareness but still
outlined plans to use vast amounts of government and commercial data on U.S. citizens to sniff out suspicious activity.

While debates rage about these two programs, though, myriad other government agencies and private companies are building
similar kinds of massive, easily searched databases on a broad range of people, all in the name of the war on terror. The
emerging systems link databases that didn't communicate previously, mixing public records, such as indictments and
prosecutions, with intelligence based largely on investigators' hunches.

The idea is to use sophisticated software to "see things that a human being can't possibly see," says Steve McCraw, the Federal
Bureau of Investigation's assistant director in the agency's office of intelligence. "We collect such voluminous amounts of data
that we need to be able to find previously unknown links, relationships and associations hidden within our own data."

But critics say the more information of varying credibility is amassed in one place, the greater the risk of overloading
investigators with irrelevant leads that cast needless suspicion on innocent people. "What we're talking about here is unverified
information and not necessarily very accurate information," says James Dempsey, executive director of the Center for
Democracy and Technology, an online policy group in Washington.

Some examples of what's in the works:

• The FBI is working with outside contractors to build an unprecedented "data mart." By importing data from other federal
agencies and linking to local police intelligence databases, the Terrorism and Intelligence Data Information Sharing Data Mart
will get instant access to a broad range of people. "Text-mining" software will then scan for common elements in more than a
billion documents from FBI field offices across the country.

• An FBI database called the Violent Gang and Terrorist Organization File is expanding rapidly. When it was launched in 1995,
VGTOF was mainly used to track violent urban street gangs. Early last year, its purpose was quietly expanded to include all
subjects of FBI domestic or international terrorist investigations. A February 2002 memo citing the 2002 Winter Olympics
shows how the FBI's definition of potential terrorists has broadened: It encompasses such categories as "anarchists," "militia,"
"white supremacist," "black extremist," "animal rights extremist," "environmental extremist," "radical Islamic extremist" and
"European origin extremist."

Because police even check VGTOF at traffic stops -- and it includes suspects with no criminal record -- inclusion in the
database is supposed to be limited to people who pose a significant threat, says Roy Weise, senior adviser to the FBI's Criminal
Justice Information Services Division. But the terrorism listings now include more than 7,000 names, alongside the tens of
thousands of gang members.

• Police intelligence files are being shared more widely. In a pilot project, several