Williams later found out he was wrongfully arrested after he was falsely identified as a shoplifting suspect in January 2020 by Detroit police after they used facial recognition to search the surveillance footage of a theft at a Shinola store. He yesterday shared the painful details with members of Congress of his nearly 30 hours in police custody, as they weigh action to rein in the artificial surveillance that is increasingly being used by law enforcement throughout the country.
His testimony highlights the costs of the technology’s expansion with little government oversight.
Republicans and Democrats have both blasted the artificial software as a threat to U.S. privacy and civil rights — one of the rare areas of bipartisan agreement in Washington. But despite this agreement and a high-profile series of hearings in the past Congress, lawmakers have taken little action to rein in the technology that has been widely used by law enforcement.
In the absence of action, facial recognition technology is being used more often and in more high-profile ways. It was notably used to identify suspects who participated in the Jan. 6 attacks on the U.S. Capitol. Six federal agencies, including the U.S. Park Police and the FBI, recently told the Government Accountability Office they had used facial recognition to find people who participated in protests after the police murder of George Floyd.
Lawmakers say it’s time to weigh facial recognition’s potential benefits against the lack of transparency from companies making products using it.
“As these trends have developed, the federal government has been largely absent,” said Rep. Sheila Jackson Lee (D-Tex.), chair of the House Judiciary Committee’s subcommittee on crime, terrorism, and homeland security.
And Rep. Jim Jordan (Ohio), the top Republican on the Judiciary Committee, expressed interest in working together. He had previously been working on developing facial recognition laws with Rep. Elijah Cummings (D-Md.), who died nearly two years ago.
“I am hopeful that we can approach this issue in a truly bipartisan matter,” Jordan said. “While we had many vigorous debates on a number of hearings, facial recognition was one area where we shared common ground.”
The hearing follows a major federal report on the expansive use of facial recognition.
Lawmakers also heard testimony from Gretta L. Goodwin, the director of homeland security and justice at the U.S. Government Accountability Office. The GAO recently issued a report calling for stricter limits on the federal government’s use of facial recognition software. The agency found its use is widespread throughout the federal government, but there are few guardrails and many agencies don’t even know what systems they’re employing.
The GAO found that 20 of 42 federal agencies relying on some kind of law enforcement use facial recognition systems. Its report said 14 of the agencies reported using the A.I. software for criminal investigations, but only one agency had a mechanism to track how employees were using such systems. Some agencies use their own databases, but others use tools from private companies such as Amazon and Clearview AI. (Amazon founder Jeff Bezos owns The Washington Post.)
The findings have added to the pressure on the federal government to increase regulation of the software, as The Technology 202 researcher Aaron Schaffer has reported. Some activists want use of the technology to be outright banned at the federal level. And activists are growing frustrated waiting for Congress.
“We don’t need facial recognition regulations, we need a full ban,” said Albert Fox Cahn, the executive director of the Surveillance Technology Oversight Project in a statement. “We can’t wait for Congress to act, so we are calling on President Biden to issue a moratorium on federal facial recognition.”
Our top tabs
A French regulator fined Google $593 million for not negotiating in good faith with publishers.
The fine is a major win for news publishers that say they should be paid fairly by Internet platforms for their content, the New York Times’s Adam Satariano reports. French regulators said Google ignored a 2020 order to negotiate a deal with publishers to use blurbs of articles in search results.
Google, which can appeal the fine, said it was “very disappointed” with the ruling. “We have acted in good faith throughout the entire process,” the company said. “The fine ignores our efforts to reach an agreement, and the reality of how news works on our platforms.”
The Justice Department tried to get records on journalists via Proofpoint.
The strategy of using Proofpoint as a means to get email records on Washington Post journalists suggests prosecutors were thinking creatively — and beyond standard email providers such as Google or Microsoft — about where to get records on journalists, Devlin Barrett and Spencer S. Hsu report. The Justice Department sought the records as prosecutors worked to identify sources for three articles written in 2017.
Proofpoint did not respond to requests for comment.
Senior Justice Department officials are working on regulations to limit how they can pursue journalists’ data when looking for sources of classified information. President Biden has condemned the practice.
Thousands of Facebook employees accessed users’ personal accounts.
Many of the engineers who improperly accessed the data used it to look up women they were interested in and, in some cases, tracked their locations, Sheera Frenkel and Cecilia Kang write in an excerpt of their new book, “An Ugly Truth,” that was published by the Telegraph. Facebook CEO Mark Zuckerberg was taken aback by the pervasiveness of the problem when then-chief security officer Alex Stamos described it in 2015.
Fifty-two employees were fired for improperly accessing user data from January 2014 to August 2015.
“We’ve always had zero tolerance for abuse and have fired every single employee ever found to be improperly accessing data,” a representative told Insider. “Since 2015, we’ve continued to strengthen our employee training, abuse detection, and prevention protocols. We’re also continuing to reduce the need for engineers to access some types of data as they work to build and support our services.”
Rant and rave
Many people weren’t surprised by the book excerpt. Former CIA analyst Aki Peritz:
Microsoft’s Chloe Condon:
Technology attorney Tiffany C. Li:
Inside the industry
Wrapping up his court testimony, Tesla CEO Elon Musk said that Tesla’s goal “is not to be a car company.”
Tensions between Musk and opposing lawyers were high for a second day as the Tesla CEO wrapped up his testimony over a $2.6 billion deal to acquire solar company SolarCity, the Wall Street Journal’s Dave Michaels and Rebecca Elliott report. Much of the questioning focused on SolarCity financial information that was given to Tesla shareholders.
Musk’s defense of Tesla’s goals beyond being a car company is part of his broader arguments that the SolarCity acquisition was necessary for Tesla in its ambitions to become a clean-energy company.
The plaintiffs in the suit argue that Musk made the deal out of his own self interest. Musk could be forced to pay as much as $2 billion to Tesla if he loses the suit. The trial is expected to last two weeks.
- President Biden nominated Deloitte consultant Alan Estevez, a former Pentagon official, to be the Commerce Department’s undersecretary for industry and security. Estevez would oversee the Bureau of Industry and Security, which works on export control and sanctions issues.
- Tech trade group ITI hosts an event on the impact of a 2020 court decision that invalidated an European Union-United States Privacy Shield today at 9 a.m.
- The Atlantic Council hosts an event on the international negotiations over digital taxes on Wednesday at 9 a.m.
- The Internet Governance Forum USA conference kicks off at 11:15 a.m. on Wednesday with a keynote by Doreen Bogdan-Martin, the director of the International Telecommunication Union’s Telecommunication Development Bureau.