September 23, 2014

Next Generation Identification

FBI Announces Biometrics Suite’s Full Operational Capability

FBI Data Center

The data center at the Criminal Justice Information Services Division in West Virginia is home to the Next Generation Identification system, or NGI.

Agencies searching the FBI’s criminal history record database for matches to their subjects are getting faster and more accurate responses—the result of the Bureau’s 10-year effort to improve its ability to provide law enforcement partners with timely, high-quality identification.

Earlier this month, the FBI announced the Next Generation Identification system, or NGI, is now at full operational capability. The system replaced the Integrated Automated Fingerprint Identification System (IAFIS), the Bureau’s longstanding repository for fingerprints. NGI’s incremental roll-out, which began in 2010, has already seen significant improvements in accuracy rates on queries, the result of new high-tech tools and algorithms that more effectively search more than 100 million records. Fingerprint matches are now better than 99 percent accurate, and hits on latent prints (prints lifted from crime scenes, for example) have tripled from 27 percent accuracy in the old IAFIS system to more than 81 percent today.

“NGI gives us this opportunity to not only upgrade and enhance technology that we’ve been using for years, but it also lets us leverage new technology that can help us do our jobs better,” said Steve Morris, assistant director of the FBI’s Criminal Justice Information Services (CJIS) Division, which runs NGI.

Enhancements under NGI include the following:

  • Repository for Individuals of Special Concern (RISC): Deployed in 2011, it’s a searchable subset of what Morris described as the database’s “worst of the worst offenders,” including terrorists and dangerous fugitives. Using a mobile device, police can take two fingerprints from a subject and remotely query the database and get immediate results. “NGI provides a quicker mobile identification for the officer,” Morris said.
  • National Palm Print System: In May 2013, NGI expanded beyond traditional finger and thumbprint capabilities to include palms. Morris said the majority of prints left at crime scenes contain hand ridges and palm prints. Just this month, one latent palm print returned a match.
  • Rap Back: Entities that conduct background checks on individuals holding positions of trust (teachers, camp counselors) can receive notifications if the individual is subsequently involved in criminal activity. Launched earlier this year, Rap Back is named for the process of reporting back when a person is involved in criminal activity.
  • Interstate Photo System (IPS): Launched this year, NGI’s facial recognition capability provides a way to search millions of mug shots or images associated with criminal identities for potential matches. Note that civil files (such as those in Rap Back) and criminal mug shots reside in a repository separated by identity group, so an innocent schoolteacher’s image isn’t going to appear when the system returns an array of possible candidates in a criminal query. “If law enforcement submits that photo, they’re going to get back possible candidates from the criminal file,” said Morris. “They’re not getting the ones from the civil file.”

In safeguarding privacy and protecting the public’s rights and civil liberties, NGI is subject to the same extensive security protections, access limitations, and quality control standards already in existence for IAFIS. A thorough privacy impact assessment is completed and submitted to DOJ for each enhancement under NGI.

The facial recognition system is not connected to the Internet or social networks or your local Department of Motor Vehicles. “Facial recognition doesn’t mean that we somehow now have this ability to go out and start collecting video feeds,” Morris said. “That’s not what this is about. It’s a technology that allows us to digitally compare criminal mug shot photos that we have in our database against one another.”

For more than 18,000 law enforcement agencies and partners—and their constituents—upgrading to NGI means increased accuracy and improved, faster intelligence. “Not only are we providing a better, more accurate technology, but we’re able to provide all these better services more efficiently,” Morris said.

Closing the Gap

The expanded capabilities of Next Generation Identification (NGI) have reduced the margin of error in identifying criminals. NGI’s biometric modalities include not only fingerprints, but also palm prints, mug shots, and scars that, when combined, can present a compelling case. Steve Morris, assistant director of CJIS, said NGI is not a new FBI authority; rather it is a tool that takes advantage of new technologies.

“We’re leveraging technology that is not only increasing the accuracy of these things, but when you combine them—a fingerprint that is 99-plus percent reliable and a facial image—you are almost providing virtual certainty.

“Right now people may say that 99 percent is pretty good. Why do we need better than that? Well, I guess that’s okay unless you’re that one guy in a 100 that the system made a wrong identification on. To be able to close that 1 percent gap—to 99.99 percent, because we’ve now added a photograph—the chances of the two systems misidentifying that person are exponentially less.”