But a number of groups, liberal and conservative, are troubled by the whole concept. They doubt such a system would be effective in identifying terrorists and say the cost in terms of loss of privacy is not worth it.
''It's not clear a computer program can spot terrorists. The idea that a computer algorithm can do this is flawed," said Jay Stanley, communications director for the Technology and Liberty Program at the American Civil Liberties Union. ''What we end up with is a totalitarian system of internal border checkpoints."
CAPPS I is already in place. It's run by the airlines and attempts to ferret out those passengers deserving greater security scrutiny by comparing their names to terrorist watch lists and analyzing such factors as whether they paid cash for their ticket, checked any baggage, or how soon before a flight they bought their ticket.
A TSA official, who declined to be identified by name, said CAPPS I is plagued by errors caused by name similarities. The official said the secondary security screenings required under CAPPS I delay travelers and waste valuable security resources.
Under CAPPS II, the official said, secondary security screenings should be triggered for about 3 to 4 percent of air travelers, which by some estimates would be close to 20 million people.
CAPPS II would require airlines to ask all passengers for their full names, addresses, phone numbers, and dates of birth. The information would be transmitted before takeoff to the TSA, which would verify the traveler's identity using unspecified commercial databases and then run the passenger's name against terrorist watch lists.
In an outline of the program released in January, the TSA said it would verify passenger identities using medical, credit, and banking databases. The agency was vague about whether the information would be used solely to identify terrorists or to identify anyone with an outstanding warrant, and it suggested that data gathered under the system could be kept for as long as 50 years.
