Does Biometric ID Verification Give Way to Data and Discrimination Risk?
Share
APEX Insight: Airlines and airports are doubling down on biometric passenger identification, but is this evolution in security practices creating data and discrimination risks in modern air travel?
Australia is seeking to overhaul its immigration process, automating entry for up to 90 percent of incoming passengers. Paper entry cards and even passport scanners would be scrapped, and passengers arriving would no longer have to interact with border control agents before picking up their luggage. ID verification would be handled with biometric scanning, via facial, iris or fingerprint recognition technology. Amsterdam’s Schiphol International Airport, in partnership with KLM, is also testing a biometric boarding process that would eschew passports, relying instead on facial recognition technology. Meanwhile, Paris Charles de Gaulle Airport is conducting trials of facial recognition software that could clear up gridlock.
Such biometric interfaces are only the outer pieces of the puzzle: Behind the scenes, passenger biometric data is passed through algorithms and existing databases to decide who’s in, who’s out and who gets a human once-over. However, the “garbage in, garbage out” rule of computer programming still applies: If a software team isn’t careful, bias can be written into the algorithm.
If a software team isn’t careful, bias can be written into the algorithm.
Data can run amok in all sorts of ways, but the stakes around biometric information are among the highest. As biometric data collection and CCTV security cameras become ever more present in modern airports, experts wonder whether these new technologies will give way to automated discrimination. Speaking about the security overhaul at Australian airports, University of Wollongong biometrics expert Katina Michael says the technology could threaten individual privacy and raises ethical quandaries that had not been properly disclosed to the public. “We are steam-training right through all of these technological transitions and we’re not really thinking about the ramifications,” she told the Guardian. “I see the perceived benefit, but what I do know is that there will be real costs, human costs, not only through the loss of staff through automation, but also through discrimination of people who may appear different.”
Clearing airport security – whether arriving at the airport or landing and clearing customs – is a slow process that’s arguably begging for disruption. However, when rolling out automated identity verification, airports will need to ensure that human bias isn’t being baked into the system.