Surveillance-tech debate intensifies as US senators seek limits on facial recognition use by cops
By Drew Harwell
Senators are taking aim at a powerful facial recognition tool and the law-enforcement purchase of Americans' personal data in a landmark bill introduced Wednesday with bipartisan support that highlights the growing debate over surveillance technologies used by federal investigators and local police.
The legislation, dubbed the Fourth Amendment Is Not For Sale Act, would ban the U.S. government and law enforcement from buying location data and other personal information without a warrant. It was introduced by Sens. Ron Wyden, D-Ore., and Rand Paul, R-Ky.
It would also block the purchase of data that was "illegitimately obtained" via an act of deception, hacking or breach of contract, a stipulation that congressional aides said would ban the police use of Clearview AI, one of the most popular facial recognition programs used by thousands of police officer across the country. The facial recognition firm pulled billions of photos from social media and other websites such as Facebook, Google and Twitter, which have accused the company of violating its rules against "scraping" personal data.
The bill represents one of Congress' most ambitious attempts yet to regulate technologies that government officials increasingly have turned to without court approval or public oversight.
Defenders of government's use of the technologies' say they provide breakthrough information in the pursuit of suspected criminals. But critics worry they represent a disturbing overreach of government power and threaten to subject Americans to an increasingly invasive surveillance regime.
"Doing business online doesn't amount to giving the government permission to track your every movement or rifle through the most personal details of your life," Wyden said in a statement. The bill, he added, "ensures that the government can't use its credit card to end-run the Fourth Amendment" rights against unreasonable searches.
Local, state and federal law-enforcement agencies have tapped private databases that gather and bundle millions of location records, pulled from a set of smartphone apps on people's phones, effectively transforming a system designed for online marketers into an evidence-gathering system for government investigation.
Getting such data from tech and cellphone companies traditionally requires a court order. But the lack of regulation has allowed police to pay data brokers for such information outright.
Investigators also have used Clearview AI to match the faces of suspects caught on camera to images pulled from social media, including in the Capitol riot investigation of Jan. 6. Traditional facial recognition systems only compare photos to official photos from driver's licenses, passports or jail mug shots.
Clearview AI chief executive Hoan Ton-That said in a statement that the company planned to "carefully review" the bill, adding, "We look forward to engaging with policymakers on the best ways to protect consumer data and continue to be a resource for law enforcement agencies." Floyd Abrams, a lawyer representing the company, said in a statement that "the Fourth Amendment does not protect downloading and analyzing photographs that people voluntarily post on the Internet, since there can be no reasonable expectation of privacy in those circumstances."
The Senate bill is co-sponsored by 17 Senate Democrats, including Majority Leader Charles E. Schumer, N.Y., and three Senate Republicans, including Paul, Mike Lee (Utah) and Steve Daines, Mont., giving it bipartisan backing that congressional aides said has given them confidence in the bill's chances to become law.
A House version was also introduced Wednesday by Judiciary Committee Chairman Jerry Nadler, D-N.Y., and House Administration Committee Chair Zoe Lofgren, D-Calif., who said in a statement, "Our digital data opens a window into the most sensitive areas of our private life, and this bill would be a major step forward in curbing surveillance abuse and protecting Americans' civil liberties."
The proposal, a Wyden aide said, would close several other surveillance loopholes revealed in the aftermath of the Edward Snowden leaks, including one allowing intelligence officials to acquire metadata about Americans' calls and emails abroad without court approval.
The bill has widespread support from civil-liberties and privacy groups that have been critical of the expansion of surveillance technologies they say represent threats to civil rights.
Susan Ariel Aaronson, director of the Digital Trade and Data Governance Hub at George Washington University, said the bill was a "good start" but did not fully address how so much personal data about Americans has been gathered and analyzed by largely unregulated companies for commercial use.
"It's been disgusting to see how U.S. agencies at the local level and intelligence agencies have misused and bought data sets to spy on us," Aaronson said. "But this doesn't get at the root problem, which is that the market for data is so opaque and we don't adequately protect people. . . . They are misusing our data to surveil us and manipulate us. That's the larger problem, and I feel like we're just dancing around it."
The bill comes amid a flurry of proposals aimed at firming up Americans' digital privacy rights. A separate proposal first reported by The Washington Post this month would ban the sale of Americans' personal data to a selection of foreign companies and government agencies deemed "unfriendly" by a U.S. government review. Another bill would restrict how U.S. companies gather and share Americans' personal information.
It also comes amid a global reckoning for artificial intelligence and other emerging technologies that have gained a pervasive foothold in everyday life.
A proposal Wednesday by the European Union's executive branch pledged to ban "high-risk" AI use, such as automated "social scoring" systems like those seen in China and indiscriminate mass surveillance. A Wyden aide said the senator's office was tracking those discussions closely but had not had any substantive conservations with EU regulators.
The Biden administration has faced growing calls to tackle the potential harms of AI software such as facial recognition, which has been cited in three reported cases of police misidentification and wrongful arrest, all of which involved Black men.
A staff attorney with the Federal Trade Commission wrote Monday that federal laws banning unfair and deceptive business practices also applied to the developers of "racially biased algorithms" and other problematic systems, noting that supposedly "neutral" technologies can still "produce troubling outcomes."
Aaronson said she found the proposed regulation's timing ironic, given the government's broader role in capitalizing on vast data sets about the general public. She pointed to a note in a document published last month by the National Intelligence Council, which outlined potential global trends over the next two decades: "Privacy and anonymity may effectively disappear by choice or government mandate, as all aspects of personal and professional lives are tracked by global networks."
The Washington Post