Post-9/11 antiterror technology: A report card

news analysis As Sept. 11 nears, News.com examines five useful ways of improving security--and five that should raise eyebrows.

Declan McCullagh
Declan McCullagh Former Senior Writer
Declan McCullagh is the chief political correspondent for CNET. You can e-mail him or follow him on Twitter as declanm. Declan previously was a reporter for Time and the Washington bureau chief for Wired and wrote the Taking Liberties section and Other People's Money column for CBS News' Web site.
11 min read

news analysis Five years after the Sept. 11 attacks on the World Trade Center and the Pentagon, the federal government's record of adopting antiterrorism technologies has been mixed.

Puffers, chemical scanners and biometrics devices are appearing in airports. Radio-frequency chips are being inserted in U.S. passports. The U.S. Army has developed machine-gun toting robots for deployment in Iraq.

But the FBI is still struggling with computer systems that are at least half a decade out of date, Homeland Security is having similar problems with inspections of shipping containers, and it's hardly clear that RFID-equipped passports are any safer from duplication by an identity thief or enterprising member of al-Qaida.

CNET News.com has compiled a list of 10 technologies, five that should be adopted more speedily to help in homeland security efforts--and five that raise at least some privacy and security concerns. Read on for the details.

In need of support
1. Going wireless: Ever since cameras on cell phones became popular a few years ago, millions of Americans have zapped grainy snapshots back and forth wirelessly. Now the chronologically backward folks at the FBI finally are entering the 21st century too.

An FBI pilot program launched last month in Washington, D.C., and New York City is designed to outfit field agents with wireless technology. They'll be able to take digital photos of a suspect, upload the images to a broadband wireless-enabled laptop, and e-mail it off to other on-the-go agents. They, in turn, can view the suspect's image--complete with that day's garb and haircut length--on a BlackBerry handheld device.

This is hardly a novel idea, of course. But it's still a useful upgrade to the FBI's existing technology, says Frederick Brink, who's in charge of the special operations division at the FBI's New York City field office.

For the hundreds of agents now trying out the mobile technology, "they don't necessarily have to use a telephone or call in, they can access this information directly in the course of their normal duties," Brink said. The FBI's central information technology outfit would like to expand the so-called "mobility pilot program" to every FBI field office, but it has not set a timetable. As for success stories from the unwired FBI set, Brink said to check back in a month or two. But for now, he said, "the feedback is already very, very positive. In general terms, the agents love this capability."

2. Better search technology: The private sector has been years ahead of the FBI not just in wireless technology, but also in search. Internet search engines have been around since Archie in 1990, followed by the original Wandex, a Web search tool developed in 1993 at MIT.

The FBI finally got a rudimentary Web-based search tool in 2004 in the form of its Investigative Data Warehouse, or IDW. It lets users (more than 13,000 people have been approved so far) to use a single Web-based front end to comb about 650 million records--ranging from intelligence wires to terrorist watch lists to no-fly lists--across multiple government agencies, including the State Department and Homeland Security. Agents say it acts as a "one-stop shop" for wide-ranging information that takes an average of three seconds to five seconds to return results.

Unfortunately, the IDW's records aren't updated in real time. Instead, the system relies on copies of documents that must be "affirmatively uploaded into the warehouse" by participating agencies, according to an 2005 auditor's report. Depending on who's in control of the data, that can happen anywhere from daily to weekly to monthly to quarterly--although in an emergency situation, updates can be sped up to hourly, FBI Chief Information Officer Zal Azmi told reporters last week.

"Right now, we don't have that Google-like search capacity to go (directly) into databases of different agencies," Azmi acknowledged. Because "timeliness of the data is critical to us in our mission," he added, a real-time "portal" is the goal, but it is "a long way from being completed...at least a couple more years."

3. Inspecting cargo containers: Could terrorists smuggle a nuclear, biological or radiological explosive device into the U.S. by hiding it in a cargo container? There's reason to think so: 11 million cargo containers arrive at U.S. seaports each year, and only a small percentage are physically inspected by Homeland Security agents (click for PDF).

U.S. Customs and Border Protection, part of Homeland Security, does have a computerized modeling system that's supposed to help identify which cargo containers should be inspected based on intelligence from sources including the CIA. It's called the Automated Targeting System, or ATS, and has been deemed a failure by government auditors in a report this year (click for PDF). They concluded that Homeland Security "has not yet put key controls in place to provide reasonable assurance that ATS is effective at targeting oceangoing cargo containers with the highest risk of containing smuggled weapons of mass destruction."

Fixing ATS would be a good first step. So would making greater use of noninvasive methods of scanning containers and preventing unions from derailing security methods. The West Coast longshoremen's union prohibits its members from driving through gamma ray scanners, even though Homeland Security officers do it routinely and the Nuclear Regulatory Commission has approved the low exposure level. Union leaders won't allow members to drive through even more promising systems using neutron-based detectors either.

4. Smarter translation software: Intelligence agencies around the world continue to face a shortage of speakers of Arabic and other languages often associated with terrorist groups. Translation can also be time-consuming.

To this end, Language Weaver has developed machine translation tools that can dynamically translate Arabic, Russian, Chinese and 10 other languages into English. In its sales presentations, the company has its software produce an English transcript of an Al-Jazeera broadcast while the broadcast is airing.

"It used to be finding a needle in a haystack; now it's trying to find a needle in a haystack in a field of haystacks," Language Weaver CEO Bryce Benjamin told CNET News.com in an earlier interview. "There is a lot of focus on getting automated tools." Language Weaver has received funding from the CIA-funded venture capital firm In-Q-Tel.

But more obscure languages like Pashtu and Somali are still unavailable for automated translations, which is why the federal government is working on its own internal projects. One of those the Defense Department's Language and Speech Exploitation Resources program, or LASER. It's designed to provide intelligence analysts and the military with speech transcription and translation capabilities. (Similar government-funded efforts are called Babylon, a portable device, and the Effective, Affordable, Reusable Speech-to-Text project.)

5. Faster chemical detection: The possibility of chemical attacks by terrorists has federal officials running scared, with some justification. The Aum Shinrikyo attack on the Tokyo subway system in 1995 using sarin gas--which killed 12 people and injured more than 5,000 people--showed that it's possible. The attack would have been deadlier if the group had been more skilled.

In open-air environments like city streets, the threat of a chemical attack is not as severe. Winds are unpredictable and, coupled with rising air currents, can quickly disperse a chemical agent unless a larger quantity is used.

But in subways, train stations and airports, the threat of a chemical attack is higher. In an article published in Time magazine in June, author Ron Suskind reported that a terrorist cell had planned a hydrogen cyanide attack on New York City subways but inexplicably called it off with just a few weeks to go.

Hazard materials teams at local police departments historically have used colorimetric tubes, which are designed to detect specific gases such as ammonia or chlorine. A pump is used to draw air samples through the tubes.

The problem, though, is that many chemicals can be used as weapons, and standard-issue colorimetric tubes will detect relatively few. "Many modern detection devices used by hazmat teams have not been thoroughly tested for their utility and reliability to detect" chemical weapons, a panel organized under the National Research Council concluded.

Detection technology, however, is advancing. The Safesite detector, for instance, can electronically determine the difference between nerve agents, blister agents, and toxic gases such as chlorine, hydrogen cyanide, and hydrogen chloride. And an article this year in the journal Analytical Chemistry describes how to use photoionization mass spectrometry to detect chemical warfare agents. That takes about 45 seconds--far speedier than the traditional way of performing mass spectrometry that can take an hour or more.

Raising privacy concerns
1. Omnipresent cameras: Soon after the Sept. 11 attacks, the number of surveillance cameras began growing even faster than the Department of Homeland Security's budget. In one of their more "alarming cases" in 2004, volunteers from the New York Civil Liberties Union counted 600 cameras in Manhattan's Chinatown alone--up from 13 in 1998.

Police claim say they're useful for fighting, if not preventing, terrorism; footage from London's extensive closed-circuit surveillance system helped to identify suspects from the July 2005 bombings on its subway system. Another argument is that cameras do double duty, helping nab drug dealers and thieves too.

Yet evidence suggests that surveillance cameras have limited use in crime prevention. For one thing, they seem to cause crime to shift to locations not near cameras: Violent crime in Britain has risen as cameras have multiplied. Some police may also be using controllable cameras to ogle women. And if face-recognition software is linked to the cameras, police can effectively compile dossiers on Americans' movements whenever they're in public places.

In Washington, D.C., the city council handed over more than $2.3 million last month for the installation of four dozen new surveillance cameras to the city's existing closed-circuit television system after a spate of 14 homicides in a two-week span in July.

It hasn't been uniformly applauded. "This is like a modern-day jail now," one resident of a newly watched apartment complex told the Washington City Paper, a local alternative newsweekly.

2. Registered traveler: Air travelers are gradually separating into a two-class hierarchy, at least for people who haven't opted out of the system in favor of flying to their destination in a small plane.

The masses sit through irksome lines at security checkpoints. But people who pay $80 a year and submit a wealth of personal information (including fingerprint and iris scans) to the government, and clear a background check conducted by the Transportation Security Administration, can sail through airport security.

This program is run by a private company called Verified Identity Pass and has been operational since July 2005 at some airports. Last week, the company announced it would expand its registered traveler program to British Airways Terminal 7 at John F. Kennedy International Airport this fall.

Melissa Ngo, an attorney at the Electronic Privacy Information Center (EPIC) in Washington, D.C., says: "Bad guys who don't have previous ties to terrorism can pass the background check and then fast-track through airport security. If certain security procedures work to reduce crime, then they should be applied to everyone, not just to those who can't or won't pay $80 per year for travel convenience."

3. Backscatter X-ray: Comic books in the 1950s promised to sell "X-Ray" specs that could see through clothing.

Now that not-terribly-accurate promise is approaching reality, thanks to a technology called backscatter X-ray. Its proponents say it's better at detecting weapons in carry-on luggage. But privacy advocates say it can show body contours that are so exact it amounts to a "virtual strip search." It's already being used in some airports.

"Keeping the radiation dose low enough to skim the skin's surface means that backscatter cannot detect weapons hidden in body folds, which would be found during a physical inspection," Ngo said. "It's unfortunate that Homeland Security money is being spent on backscatter even though the government complains it doesn't have enough money to screen all carry-on and checked baggage and air cargo."

The best way might be to let passengers decide: Some airlines could use backscatter X-ray technology if they chose, and some would use pat-down techniques. But instead, the TSA and local governments tend to set one-size-fits-all rules. (For its part, the TSA says backscatter technology is being used with a privacy algorithm to "eliminate much of the detail shown in the images of the individual while still being effective.")

4. "Brain fingerprinting": Lawrence Farwell invented what he calls "brain fingerprinting," which tries to measure whether the mind recognizes familiar stimuli such as words or photographs.

It relies on the discovery that an electrical signal known as P300 tends to be emitted from a brain about three-tenths of a second after it recognizes a familiar stimulus. The idea is that a murderer's brain will emit P300 if he's shown the victim's face or the crime scene. (The CIA gave Farwell about $1 million in research expenses.)

Farwell created a company called Brain Fingerprinting Laboratories to commercialize the idea and has met with some success in law enforcement circles. An article this year on Officer.com says the "technology has the potential to be applicable in an overwhelming number of cases."

One judge in Iowa has ruled that the technique is admissible. The Iowa Supreme Court subsequently said in 2003 that Farwell's testing of the brain of Terry Harrington, a convicted murderer, showed "that Harrington's brain did not contain information about (the) murder. On the other hand, Farwell testified, testing did confirm that (the murderer's) brain contained information consistent with his alibi." The Supreme Court granted Harrington a new trial--based on the fact that the police withheld evidence, not because of the brain fingerprinting.

Does it really work? FBI agents who worked with Farwell think so. At least one judge was sufficiently credulous.

But a government report includes an important caution from J. Peter Rosenfeld of Northwestern University's psychology department who has done extensive research into P300. First, Rosenfeld says, there has been a lack of peer-reviewed studies. The report adds: "Rosenfeld does not believe that the developer had done the extensive validation of the test items for field use...Rosenfeld questioned the developer's claim of a 100 percent accuracy rate. For example, he raised concerns regarding whether the developer omitted inconclusive results from the totals."

5. DNA dragnets: In the last few years, as DNA testing kits have become cheaper, police have begun to engage in widescale testing of criminal suspects.

In one case scheduled to be heard by the 5th U.S. Circuit Court of Appeals on Thursday, police in Baton Rouge, La., demanded that 1,200 men provide DNA samples without seeking a court order. Shannon Kohler refused to provide a DNA sample. Police retaliated by naming Kohler as a suspect in a rape-murder case and sought a search warrant. He was eventually cleared.

Kohler is not alone. When a DNA dragnet is set up, police tend to view anyone who won't voluntarily participate as a suspect. It's also not clear what happens to the DNA sample. Will it be destroyed when the investigation is over or kept on file forever?

DNA dragnets have nabbed the wrong suspects before. In Miami, a man was incorrectly charged with rape. In Kansas, the "BTK Killer" was located through traditional police work--even though 1,300 men were tested in the DNA dragnet. Their DNA samples were all kept on file.

CNET News.com's Anne Broache and Michael Kanellos contributed to this report.