What you need to know about facial recognition at airports

0

As Americans feel more comfortable traveling during the pandemic, international travelers may find a new ID system used by US Customs and Border Protection (CBP) when they return home in the form of biometric facial recognition. Following a directive from the Congressional 9/11 Commission to increase border security using biometrics, CBP began bolstering the technology in 2018 under a program called Simplified Arrival. Among other biometric measures available, including iris scans and fingerprints, the agency selected facial recognition, which uses a computer algorithm to match a photo taken in person at airport immigration or at another border control point to the traveller’s passport or visa photo.

“We automated a manual process,” said Diane Sabatino, CBP’s deputy executive deputy commissioner, who oversees the biometrics program.

Some privacy advocates have questioned the use of technology. Addressing fairness, Senators Roy Blunt, Republican of Missouri, and Jeff Merkley, Democrat of Oregon, sent a letter to the agency in January requesting more information “to ensure that reported individuals are treated in a fair manner.” safe, fair and non-invasive given the imperfection of facial recognition software.

The following are excerpts from an interview about the issues with Ms. Sabatino, edited for length and clarity.

Credit…CCP

When we looked at different biometric technologies – fingerprints, iris and facial scans – we landed with the face because it’s such a simple process. Travelers show up with their documentation and pose for a quick photo in seconds. The agent has the data they need based on a discussion with the traveler about the purpose of the trip and can ultimately decide if further investigation is needed. We can now take advantage of more efficient technology to make comparisons. The officer remains the ultimate decision maker. Travelers can opt out.

It’s a simplified process. One of the benefits is to help agents determine travel intent more efficiently. It’s also better for identifying potential imposters. And the third element that we hadn’t considered was the additional health benefits. We have improved security at a time and place where individuals are already expected to show up for identity verification, and now we are adding contactless travel and limiting the spread of pathogens. It wasn’t something we envisioned when we developed it, but it certainly made sense.

The manual check takes 10 to 30 seconds, depending on environmental factors. Someone outside at a land boundary might be more difficult due to lighting. As we automate and refine facial recognition technology, we take two to three seconds to verify the match. The match is a tool in the whole process. This tool does not make the decision to admit or require further investigation. It’s the officer and the set of circumstances. The priority is safety.

Since the rollout, over the first three years or so, primarily in the air passenger environment and a little in the maritime, we have identified around 300 imposters using the technology. That doesn’t mean we wouldn’t have identified them otherwise. Last year, at land pedestrian crossings on the southern land border, he caught about 1,000 to 1,100.

Our business use case is to identify individuals at a time and place where they would normally expect to show up for identity verification. We do not grab images or scrape social media. People present passports and we have a repository to draw from and create galleries before they arrive using US passport photos and photos of those who have applied for visas. We therefore build these galleries in the airport and maritime environment on the basis of the information already provided for identity verification. We compare it to the information we have.

And we make sure there is secure encryption. When a gallery is created, that photo is not attached to any information and cannot be reverse-engineered to be compromised. The design is based on the privacy measures we knew we had to put in place. Images for US citizens are retained for less than 12 hours and often much less.

It is certainly something that we are very attentive to. We have partnered with the National Institute of Standards and Technology to provide information about the program. Our high performing algorithms show virtually no demonstrable difference in demographics.

We post signs at all entry points. Persons withdrawing must notify the officer during the inspection. It would then revert to the manual process.

We have deployed it in pedestrian routes at land borders. In the air environment, we cover around 99% with simplified arrival. The land border is the ultimate border. We just completed a 120-day pilot project in car lanes in Hidalgo, TX and will be evaluating the outcome. At the cruise terminals, we are in the 90% range. We work with nine major carriers at eight ports of entry, including Miami, Port Canaveral and Port Everglades, all in Florida.

We welcome the scrutiny from privacy groups. We want to be able to tell and share the story of the investment we’ve made in privacy. There are so many myths and so many misinformation confusing what we do with surveillance. Whenever new technology is deployed, there are always legitimate concerns. We welcome these questions. They help us respond better when we build these systems.

Elaine Glusac writes the Frugal Traveler column. Follow her on Instagram @eglusac.

Share.

Comments are closed.