| Big Brother May Be Watching, but for How Long? |
 | | A surveillance camera that is part of a Detroit police program using facial recognition technology. Brittany Greeson for The New York Times | |
| Each week, we review the week's news, offering analysis about the most important developments in the tech industry. Want this newsletter in your inbox? Sign up here. |
| Hi, I'm Jamie Condliffe. Greetings from London. Here's a look at this past week's tech news: |
| Reports that Immigration and Customs Enforcement analyzed millions of motorists' photos using facial recognition without their knowledge outraged privacy advocates. And an outcry erupted in Detroit over the use of the technology on data from the city's prolific closed-circuit television system. |
| Proponents of facial recognition by government and law enforcement argue that it's an important security tool. But critics keep getting louder and more numerous. They contend that facial recognition is a fundamental invasion of privacy and that it is inaccurate and often biased against minorities. They also argue that the longer it's used, the more entrenched it will become, making it harder to stop in the future. |
| "I think this has created a perfect storm for people to realize that substantive change is required," said Evan Selinger, a philosophy professor at the Rochester Institute of Technology. Mr. Selinger, along with his colleague Woodrow Hartzog, last year called for an outright ban on the use of the technology by government and law enforcement. On Tuesday, the digital rights group Fight for the Future announced its own push for an outright ban. |
| But how realistic are such proposals? |
| There's some wind at their backs. Many lawmakers — from Representative Alexandria Ocasio-Cortez, Democrat of New York, to Representative Jim Jordan, Republican of Ohio — think something must change. |
| "It's time for a timeout," Mr. Jordan said Tuesday on National Public Radio. "Let's figure out how we can put in safeguards that protect our fundamental liberties." |
| But in a House hearing on Wednesday examining the Department of Homeland Security's use of facial recognition, some lawmakers rejected the idea of a ban. It would be "an easy way to avoid hard questions," said Representative Mike Rogers, Republican of Alabama. |
| Judith Donath of the Berkman Center for Internet and Society at Harvard said arguments about security were "likely to override" concerns about the technology. And "Congress has historically been reticent to influence state and local policing," said Alvaro Bedoya, director of the Center on Privacy and Technology at Georgetown University Law Center. |
| If no full ban, though, then what? |
| Three routes stand out. First, a moratorium, like the one that civil rights advocates sought from the House Committee on Homeland Security on Tuesday, when they asked for the Department of Homeland Security to "immediately suspend" use of facial recognition. |
| Then there's a growing trend for local governments to introduce bans: San Francisco was the first city to restrict use of facial recognition by city authorities, and then Somerville, Mass., did something similar. A quasi-national ban could bubble up if many more follow suit. |
| Professor Donath argues that it may be easier to regulate through limits on how facial recognition data can be collected, stored and used by law enforcement and government. |
| It looks increasingly as if Big Brother's gaze is destined to be limited. The questions now appear to be how, and by how much. |
| Are we thinking about 5G all wrong? |
| Stop me if you've heard this before: We're losing the 5G race to China, Huawei's dominance in the technology around the globe is a threat to national security, and we had better clamp down on it. |
| But Tom Wheeler, a former chairman of the Federal Communications Commission and now a visiting fellow at the Brookings Institution, explained that things were more nuanced. It's a wide-ranging (often technical) analysis of the broader 5G policy landscape, and it's worth a read if you have the time. |
| I chatted with him about it, and here are a few of his points about the China situation that I thought were interesting: |
| ■ Coming first may not be important. President Trump has said that "the race to 5G is a race that we must win," but it's not clear what that means. What might winning look like? "The U.S. wasn't first in any of the Gs, and yet it's the dominant force in the wireless ecosystems," Mr. Wheeler said. "I believe that winning is dominance over the things that use the network," he added, as opposed to simply having the network in place first. |
| ■ Railing against Huawei could undermine the United States' dominance. The Trump administration has been trying to cut off Huawei from American tech, most recently by blocking American companies from selling to it. (That has softened a little, but the effect is unclear.) "My concern is that these trade policies are forcing the Chinese to develop their own alternatives" to United States technology, Mr. Wheeler said. The danger: that the resulting products and services are not just good but made widely available, and cheaply, too. "That will have an impact on our ability to lead," Mr. Wheeler said. |
| ■ So, now what? "There is a need for a policy strategy," Mr. Wheeler said. "We've got to define winning, and then do it!" |
| Killing software is hard |
| Imagine you wrote some software that you thought was kind of naughty but also interesting, and you then made it publicly available. Then people were like: Sure, interesting, but not naughty — intrusive, distasteful and dangerous. |
| So you decide to take it down! |
| Only, people who had downloaded it then uploaded it, putting it on code-hosting websites like GitHub, which is owned by Microsoft. Policing what is on those code-hosting sites is nearly impossible. Once software is published online, there's no getting it back, really, because there are so many corners of the web to put it in. |
| Anyway, The Verge wrote: |
| "GitHub is banning code from DeepNude, the app that used A.I. to create fake nude pictures of women. Motherboard, which first reported on DeepNude last month, confirmed that the Microsoft-owned software development platform won't allow DeepNude projects. GitHub told Motherboard that the code violated its rules against "sexually obscene content," and it's removed multiple repositories." |
| So, the horrors of nudifying A.I. are on the loose. Forever. |
| Some stories you shouldn't miss |
| ■ Twitter backed off a broad overhaul of its policies on speech, focusing on removing hate against religion instead of all "dehumanizing" content. |
| ■ Mr. Trump hosted a social media summit. Facebook and Twitter weren't invited, but a long line of pro-Trump influencers were. How democratic! |
| ■ Huawei's state ties remain under scrutiny. A study of leaked résumés said there was "strong evidence" of "a deep and lasting relationship between Huawei, its employees and the Chinese state." |
| ■ The Pentagon's cloud contract was hit by more controversy. The Wall Street Journal reported on a 2017 meeting between Jim Mattis, who was the defense secretary, and an Amazon executive. |
| ■ Zoom, a videoconferencing company, brushed off concerns about the security of its software for Macs, only to fix the issues after continued outcry. |
| ■ France's digital tax is in Mr. Trump's cross hairs. His administration will investigate whether it amounts to an unfair singling out of United States companies. |
| ■ Instagram introduced new anti-bullying features, including A.I. that urges users to reconsider sending potentially offensive messages. |
| ■ The House antitrust subcommittee is hauling in Big Tech. It asked executives from Google, Amazon, Facebook and Apple to testify at a hearing on Tuesday. |
| ■ Alexa is going to deliver health advice in Britain. What could go wrong? |
| ■ Oh, yeah: Google workers reportedly listen to its Assistant recordings, just as Amazon's listen to Alexa's. |
No comments:
Post a Comment