Legislative Eye Turns to Growing Facial Recognition Field
As billions of dollars are invested in facial recognition and companies look for contactless solutions, new legislation is addressing privacy and bias concerns.
As the use of facial recognition technologies expands, so do the concerns about privacy, consent and baked-in algorithmic bias — all of which mean that the potential harms of the technology are more likely to fall upon marginalized communities. Now legislation addressing the shortcomings of the field is coming just as its enterprise potential amid COVID-19 is being explored.
Last month, Portland, Oregon, became the first U.S. city to put forward legislation to ban the use of facial recognition by private businesses and public spaces. The Oregon city’s council will vote on two ordinances about the use of facial recognition technology — one that outlaws the use of the technology by the city’s government bureaus, and another that bans its use in places of public accommodation by private entities — on Aug. 13.
If passed, Portland’s legislation could become the inspiration for similar restrictions on facial recognition technology in other jurisdictions — which could hamper development in the sector, which Fortune Business Insights predicts will be worth $7 billion by 2024 and nearly $13 billion globally by 2027. Major tech firms such as Google, Microsoft, Amazon, Facebook and Apple are players in the field, but smaller firms, academia and Chinese companies are also investing in research and development in facial recognition.
Modern facial recognition systems use biometrics to map facial features from an image, then uses a variety of algorithms to compare that data with a database of known faces — for example, a collection of drivers’ license photos. Advancements in deep learning and 3D modeling have made facial recognition systems increasingly accurate, but none is perfect.
Significantly, there is much evidence of algorithmic biases that affect the accuracy of such systems in ways that exacerbate existing societal inequalities. “A study from the National Institute of Standards and Technology concluded that the chances of a false positive in facial recognition technology are 100 times higher in Black and Asian people than white people,” said Hiral Atha, CEO of MoveoApps. That, along with privacy concerns, is behind much of the push to regulate the use of the technology.
Risks and Benefits
If passed, the now-draft legislation in Portland would ban the use of facial recognition technology in Airbnb rentals, banks, entertainment venues, homeless shelters, public transit stations, restaurants, senior centers, stores and service businesses such as doctor’s offices.
However, airports, places of worship, public schools, private clubs, private homes and non-public areas of private workplaces are not included in the draft legislation.
The draft Portland law controlling the use of facial recognition technology by local government includes law enforcement. If the law passes, it would add the city to the growing list of those preventing police from using the technology. Boston, Oakland and San Francisco have already introduced bans on the use of facial recognition by government and police forces.
Facial recognition laws could be just the beginning of legislation addressing the use of surveillance technologies in Portland. Future laws could encompass biometric, location, visual or voice data, OneZero reported.
As the discussion around the use of facial recognition technology unfolds, consent is key, said Sanjay Gupta, vice president and global head of products and corporate development at Mitek.
“Most people would agree that using biometrics for the purposes of mass public surveillance is problematic,” Gupta said. “However, when consent is given by users, the benefits to security and digital commerce are enormous.”
There are some uses of the technology that benefit consumers, Gupta pointed out — for example, the use of facial recognition for digital and contactless commerce. These uses are increasingly appealing right now, as businesses invest in both online operations and contactless payment options and look for ways to enforce social distancing in public places. For example, the Wall Street Journal reported that several professional sports teams are considering facial recognition as a means of contactless entry to stadiums when fans can return to watch games in person.
Other applications of the technology have societal benefits, Atha said — for example, the use of facial recognition to help find missing children or runaway subjects.
In other parts of the world, the use of facial recognition technology is viewed differently. In Singapore, to give one example, Universal Studios visitors must pass through facial recognition scanners to enter the park, and DBS banking customers can use it to log into their bank accounts via their phones. China is already a leader in the facial recognition field, with Megvii among the country’s leading AI startups and Alibaba investing billions in the technology.
And many North Americans also encounter and use facial recognition technology every day. In late July, a Reuters investigation revealed that Rite Aid installed facial recognition in some of its stores — mostly those in predominantly low-income, non-white neighborhoods in New York City and Los Angeles — over a period of eight years. Also in July, Facebook settled a lawsuit about the unauthorized collection and storage of biometric information from millions of users for $650 million. And multiple times a day, millions of people use their faces to unlock their iPhones.
Is R&D the Solution?
Given both the already widespread use of facial recognition technology and its potential benefits, blanket bans on its use aren’t the solution to issues like bias or privacy concerns, Atha said.
“A ban only leads to unethical companies with low reputation risk abusing the technology by shady means,” Atha said. “Proper regulation would mean high acumen tech companies will produce better technology while minimizing the pitfalls.”
The biometrics market is expected to take a hit amid the COVID-19 pandemic, but facial recognition is an increasing part of that market. Investment in research and development with an eye to making facial recognition both more accurate and more inclusive would reduce bias and false positives, Atha said.
And the ethical use of facial recognition technology has business value, and that will continue to drive R&D investment in the field, Gupta said. “Legislative moves like that being discussed in Portland shouldn’t bring innovation to a halt,” he said.
Illinois’ Biometric Information Privacy Act, which requires specific consent for the use of biometric data, is a useful parallel, Gupta said. It’s the legislation Facebook ran afoul of.
“While there is a requirement for consumers to ‘opt in,’ this hasn’t diminished the value these technologies provide to businesses in the state,” he said. “That experience shows that even with explicit opt-in and consumer consent requirements, biometrics can play an important role in pushing safe, frictionless technology experiences forward.”
About the Author
You May Also Like