Microsoft President Brad Smith tells me he is determined to get other technology companies on board with a federal law to put limits on the use of controversial facial recognition technology. That’s a stark departure from the typical Silicon Valley playbook, which is usually to ask for forgiveness rather than permission.
“The time has come for people to have an opinion,” Smith said in an interview after an event yesterday unveiling Microsoft’s recommendations about what should be in such a law. “I’m optimistic that the companies that haven’t yet defined a point of view will do so soon.”
Smith says the industry needs guardrails in place as early as next year to limit potential abuses of the technology for surveillance or discrimination. His bold proposals include limiting the government’s use of facial recognition for surveillance without a warrant and requiring unbiased, third-party assessments of the technology to determine whether it’s accurate or biased. He has been meeting with members of Congress to talk about these ideas.
But getting the rest of Big Tech on board with a federal law will not be an easy task. Technology companies are chasing lucrative contracts– especially from law enforcement, the military and government agencies — and these kinds of limits could potentially complicate those deals. Companies may also not want to deal with extra red tape as they refine the nascent technology.
So far, other companies have mostly remained silent. Amazon, a major provider of the technology, for instance, hasn’t publicly weighed in on where it stands on Microsoft’s proposals. An Amazon representative declined to comment about whether it agrees it’s time for facial recognition regulation. (Amazon founder and chief executive Jeffrey P. Bezos owns The Post.)
Microsoft’s support for comprehensive regulation is a major break from how technology companies have previously dealt with Washington — often only coming to the table under pressure of a major scandal. Michael Posner, a business ethics professor at New York University, says in recent years there’s been “a souring and a loss of trust writ large,” in the American public’s attitudes toward Big Tech. “This represents an effort to address that gap in trust and begin to demonstrate what technology companies need to do to be responsible public citizens.”
Here are a few key things Microsoft wants to see in a facial recognition law:
- Law enforcement surveillance of people in public spaces should be limited, and used only if a court order has been obtained. Exceptions should be considered for emergencies that could result in death.
- People should review the results of facial recognition in high-stakes scenarios, such as when it could harm a person or restrict their movements.
- Companies using facial recognition should still have to comply with anti–discrimination laws.
- Companies should be required to be transparent about the limitations and capabilities of facial recognition technology.
Microsoft’s embrace of regulation threads the needle in Washington at a time when Google, Facebook and Amazon are facing an unprecedented amount of scrutiny. As executives such as Google’s Sundar Pichai and Facebook’s Mark Zuckerberg are summoned in front of congressional committees — and Trump attacks Amazon on Twitter — Microsoft has taken great pains in recent years to position itself as a collaborator with policymakers on issues from election security to privacy.
Even so, Smith said Microsoft won’t be exempt from regulatory action taken against the technology industry at large. “When Congress acted in the 1930s to regulate the banks, they didn’t create an exception for the banks they liked,” he said. “There’s a powerful incentive for everyone to focus on how we get the answers right on the future of technology and how it affects society here and around the world.”
To be sure, Microsoft also has a strong business incentive to ensure other companies buy into its principles. As it unilaterally adopts more stringent principles for its use of facial recognition, it’s a competitive risk. Smith tells me the company is losing deals because of its tough stance on how its tech should be used. The company in recent months passed on a deal with law enforcement because it felt its facial recognition technology was not yet ready for way it was going to be used, he said.
“It would be used in a way where the risk of error would be too high,” Smith said. Microsoft has also passed on deals where the technology would be used for domestic surveillance in other countries. “In that kind of situation, we just did not feel comfortable about the human rights dynamics involved.”
Microsoft believes industry buy-in is needed to avoid a “race to the bottom,” but other companies who have not taken such a public stance may still be in the running for those contracts.
Amazon works with law enforcement on facial recognition, and it has marketed its system to police departments as a way to identify criminals. It is deployed for officers’ use in Oregon and Florida, my colleague Drew Harwell reported earlier this year, and the company pitched its system to Immigration and Customs Enforcement officials earlier this year. Amazon’s technology has prompted a slew of backlash from privacy advocates — and some lawmakers such as those in the Congressional Black Caucus who fear the software could misidentify people of color and have “profound negative unintended consequences… for African Americans, undocumented immigrants, and protesters.”
Amazon has defended the work it does with law enforcement. “[M]achine learning is a very valuable tool to help law enforcement agencies, and while being concerned it’s applied correctly, we should not throw away the oven because the temperature could be set wrong and burn the pizza,” the company wrote in a blog post this summer.
Microsoft came under fire over the summer as employees raised the concern that its own technology could be used by ICE per a cloud computing agreement the company has with the agency, Drew also reported. But the company later said it doesn’t work with ICE on facial recognition. Microsoft took that stance as it first unveiled its push for facial recognition legislation this summer.
BITS: Tech executives took a step toward mending their tense relationship with the Trump administration at a White House meeting yesterday, where they discussed tightening collaboration with the government on artificial intelligence and addressed other issues such as 5G technology and quantum computing, according to the Wall Street Journal’s Douglas MacMillan. Ivanka Trump, President Trump’s adviser and daughter, helped organize the meeting and Trump himself “briefly joined” the roundtable, MacMillan reported. The gathering included Google chief executive Sundar Pichai, Microsoft chief executive Satya Nadella, IBM chief executive Ginni Rometty, Oracle chief executive Safra Catz and Qualcomm chief executive Steve Mollenkopf.
Pichai, who is set to testify before the House Judiciary Committee next week, said in a statement that the discussion was “productive and engaging,” the Associated Press’s Matt O’Brien reported. But executives from other top technology companies that have been in the political spotlight lately didn’t make the guest list. MacMillan reported Apple, Amazon and Facebook weren’t invited — despite their dominance in artificial intelligence.
NIBBLES: The release of a new Apple Watch health app shows that “the revolution in medicine and wearable tech is not yet here,” The Washington Post’s Christopher Rowland reported. That’s because the new app, which aims to detect atrial fibrillation, is not intended for use for people who actually have that condition. “The latest gadget worn on the wrist is simply not accurate enough to handle the task of assessing serious medical conditions, according to health officials,” Christopher wrote. “It’s mostly a gateway for conversations with your doctor, and Apple’s detailed setup screens for the new watch are loaded with warnings and explanations, including one that informs people with atrial fibrillation that the Apple Watch app is not for them.”
Nevertheless, the Apple Watch does have the potential to help improve health care. “There are possibilities for future advances, such as gathering electrical signals from the heart in real time, at the precise moment when a patient feels ill or a fluttering heart,” my colleague wrote. Ben Scirica, a cardiovascular specialist at Brigham and Women’s Hospital in Boston, noted: “Storing that data and sending it to a doctor will lead to better options for patients and doctors.”
BYTES: Lyft said it confidentially filed a draft registration statement with the Securities and Exchange Commission for an initial public offering — and in doing so positioned itself to beat Uber in their race to become the first publicly traded ride-hailing company. Lyft, which is a significantly smaller company than its rival, could go public in March or April while Uber aims to do so during the second half of 2019, the Wall Street Journal’s Maureen Farrell and Micah Maidenberg reported. They also noted that the “IPO will be a test” of the value that public investors place on ride-hailing companies. “Uber, Lyft and a host of other ride-hailing firms have received vast amounts of money from private investors at high valuations but still need a lot more capital as they continue to generate big losses.”
There would be several advantages for Lyft in hitting the public market before Uber. “If you’re an institutional investor who is eager — after years of watching two meteoric privately held startups amass value — to get exposure to U.S. ride-hailing, there will be a few months during which you will have one option and one option only: Lyft,” Recode’s Theodore Schleifer wrote. Additionally, it would also help Lyft build name recognition. “IPOs aren’t as critical anymore for raising money (startups can do that easily thanks to venture capitalists),” according to Schleifer. “But they are critical moments for marketing — and Lyft now has that thunder.”
— Uber is getting ready to resume testing its driverless cars, but on a smaller scale than the company did before it paused its efforts following a fatal crash in March. “Starting within a few weeks, it plans to run the vehicles on a mile loop between two company offices in Pittsburgh,” the New York Times’s Daisuke Wakabayashi and Kate Conger reported. “They won’t operate at night or in wet weather, and they won’t exceed 25 m.p.h., Uber said Wednesday.”
Still, “considerable issues” remain with the company’s autonomous vehicle technology, according to the Times. For instance, Wakabayashi and Conger reported that “as recently as a few weeks ago, the company’s autonomous vehicle unit, Uber Advanced Technologies Group, or A.T.G., was still experiencing track testing ‘failures’ on different versions of its software, according internal company emails.”
— Facebook plans to extend its political ads transparency efforts to India as the country heads toward general elections in 2019. “Facebook said Thursday that it will also maintain an online searchable Ad Library, as it has in other markets, which will document all the ads related to politics from a particular advertiser alongside other information such as range of impressions, demographics that saw the ad, and the budget that went behind an individual ad,” Manish Singh reported for VentureBeat. The move is a reminder that Facebook’s election security problems aren’t limited to the United States.
— More technology news from the private sector:
— U.S. tech investors have a message for France’s president: Take it easy on taxes. “President Emmanuel Macron must pursue tax-friendly policies towards foreign investors if he wants French startups to flourish, visiting Silicon Valley investors told Reuters on Thursday,” Reuters’s Mathieu Rosemain reported. More on the context of the visit, per Reuters: “The visit coincided with a threat by France to tax digital giants at a national level from next year if European Union states cannot reach an agreement on a tax on digital revenues.”
— More technology news from the public sector:
— Recent staff changes at the augmented reality start-up Magic Leap suggest a lack of diversity in the top ranks of the company, TechCrunch’s Lucas Matney reported. Brenda Freeman, chief marketing officer, and Rachna Bhasin, chief business officer, are taking advisory positions at Magic Leap, TechCrunch reported. “With the two executives, both women of color, taking a step back from the startup’s C-suite, there seems to be a strong lack of diversity among the startup’s top executives, a group that includes a chief games wizard and chief futurist but does not appear to have a single female chief officer,” Matney wrote.
— Terrell McSweeny, a former commissioner at the Federal Trade Commission, is joining Georgetown University Law Center’s Institute for Technology Law and Policy as a distinguished fellow.
— Dane Butswinkas, a Washingtion trial lawyer, is joining Tesla as its general counsel, Reuters reported.
— Today in funding news:
- Google chief executive Sundar Pichai testifies before the House Judiciary Committee on Dec. 11.
- House Armed Services subcommittee hearing on the Defense Department’s “artificial intelligence structure, investments, and applications” on Dec. 11.
- House Energy and Commerce subcommittee hearing titled “RAY BAUM’S Act: A bipartisan foundation for bridging the digital divide” on Dec. 11.
- The Washington Post holds a Technology 202 Live event on technology, mobility and the future of cities on Dec. 13.
The Trump administration struggles to explain its own trade agreement:
From “Black Panther” to “Pose,” highlights from the Golden Globe nominations:
10 times George H.W. Bush’s humor brought laughter to his loved ones in mourning: