Dave Gershgorn | The Verge

DuckDuckGo launches new Email Protection service to remove trackers

2021-07-20T08:30:00-04:00

DuckDuckGo is launching a new email privacy service meant to stop ad companies from spying on your inbox.

The company’s new Email Protection feature gives users a free “@duck.com” email address, which will forward emails to your regular inbox after analyzing their contents for trackers and stripping any away. DuckDuckGo is also extending this feature with unique, disposable forwarding addresses, which can be generated easily in DuckDuckGo’s mobile browser or through desktop browser extensions.

The personal DuckDuckGo email is meant to be given out to friends and contacts you know, while the disposable addresses are better served when signing up for free trials, newsletters, or anywhere you suspect might sell your email address. If the email address is compromised, you can easily deactivate it.

These tools are similar to anti-tracking features implemented by Apple in iOS 14 and iOS 15, but DuckDuckGo’s approach integrates into iOS, Android, and all major web browsers. DuckDuckGo will also make it easier to spin up disposable email addresses on the fly, for newsletters or anywhere you might share your email.

Tackling email privacy has been a major goal for DuckDuckGo, as the company pushes for privacy-friendly methods for various online tasks. The company began with its eponymous DuckDuckGo search engine and has more recently introduced its own mobile browser and desktop browser extensions to remove trackers while surfing the web.

Email trackers exist in more than 70 percent of mailing lists, according to a highly cited 2017 study. Once implemented, they let advertisers figure out when you open email, where you are when you open it, and which device you’re using. Removing trackers from email removes data points from lists building hidden advertising profiles on you, which have become a priority for privacy advocates in recent years.

Weinberg said that consumer research showed that asking people to switch to an entirely new email address and provider would be a tough pitch. So instead of creating a new email service, the Email Protection tool acts as an intermediary layer guarding access to your inbox.

“Our high-level goal is to make simple privacy protection available for everyone anywhere they are,” CEO Gabriel Weinberg tells The Verge. “So we’re trying to build this into any major browser or operating system or email provider you use.”

iPhone users might recognize the concept of a relay email address, a feature that Apple introduced with iOS 14 called Hide My Email. When signing up for an app through an Apple device, your iPhone can suggest a randomized Apple email address. Emails hit the random address, preventing the app from knowing your real email. In Apple’s iOS 15, the company is introducing similar features to confuse email trackers and protect privacy.

Weinberg explains that DuckDuckGo’s Email Protection is different than Apple’s offerings because it’s cross-platform, meaning you have a unified experience with disposable email addresses on mobile, desktop, iPhone, or Android. But it also differs in how the software zags on trackers. Apple loads trackers on its own servers, which send erroneous information back to the trackers’ servers. DuckDuckGo just removes them from the email before they are ever loaded. This difference to most users is negligible, but it is indeed a difference.

Setting up the tool is straightforward. Users sign up through DuckDuckGo’s mobile app, by going to the Settings page and tapping on Email Protection. They can then join the waitlist, which DuckDuckGo says is novel because it doesn’t collect an email address — it stores a timestamp in the app which is used to designate the person’s order in line.

DuckDuckGo estimates the wait will be a couple of weeks, after which users will get a notification to set up the feature. The setup process includes a bit of light reading about the privacy features and choosing a new “Duck.com” email address. One feature that the process stresses is that DuckDuckGo doesn’t see or save your email; the tracker removal is done in its servers’ memory (or RAM) and not ever written to a disk or hard drive. DuckDuckGo also built the forwarding software from scratch, so it didn’t have to rely on a third-party also processing the emails before they hit your inbox.

The Email Protection feature also integrates with the DuckDuckGo mobile browser and desktop browser extension, giving people the option to fill in their “@duck.com” address or generate a disposable address that forwards email to their inbox. The mobile and desktop browser extensions are linked by opening a link on the Email Protection welcome email in your desktop browser.

From then on, when email is received, scanned, cleaned of trackers, and forwarded to your email, DuckDuckGo inserts a small bar at the top of the email notifying of any trackers removed. Clicking on that bar allows a user to get more information on the trackers blocked or deactivate a disposable address that has been compromised.

Crucially, while a user can respond to an email they receive on a “@duck.com” address, it can’t be used to initially send an email, reducing the tool’s usefulness for laundering harassment.

The feature is launching in a beta test, and DuckDuckGo is still exploring how it will function in the real world. The company is also working on a privacy-focused desktop browser, which it expects to finish by the end of 2021, part of a broader push to offer privacy-focused alternatives for every major activity online.

“Wherever you go online, we want to protect you,” Weinberg said.

Trump supporters charged in plot to bomb Facebook and Twitter headquarters, among other targets

2021-07-16T16:00:01-04:00

Two California residents have been indicted for allegedly planning to firebomb the offices of California’s Democratic governor in Sacramento after the 2020 election, and had also discussed bombing Facebook and Twitter, federal court records show.

The pair, who were indicted on conspiracy charges as well as illegal possession of firearms and explosives, were part of an anti-government militia with a strong pro-Trump agenda called the Three Percenters. The conspirators’ texts alluded that their planned actions would send a message to tech companies like Facebook and Twitter days after they banned Trump from their platforms.

Texts recovered by the FBI indicated that the two men believed President Trump won the 2020 election, and that they wanted to go to “war” to keep him in office.

The first man arrested, Ian Benjamin Rogers, had been caught on January 15th, just days after allegedly texting with the other conspirator about plans for the bombing. The FBI found nearly 50 firearms, some illegal, and five pipe bombs in a search of Rogers’ home and business after the arrest. The second conspirator, Jarrod Copeland, was arrested on July 15th.

While allegedly planning to bomb Democratic buildings, the two men also allegedly discussed attacks against Facebook and Twitter headquarters, a motive that the FBI considers in its lawsuit against Rogers to be retaliation for the platforms banning Trump days earlier.

“We can attack Twitter and democrats easy right now burn they’re [sic] shit down,” Rogers wrote, according to the FBI.

The records also show that he texted “I’m thinking sac office first target,” which the FBI believes is an allusion to California Governor Gavin Newsom’s office. “Then maybe bird and face offices.”

“Bird” and “face” offices are fairly clear stand-ins for Twitter and Facebook, as Twitter’s logo is a bird and both companies had recently banned Trump from their platforms.

“Sad it’s come to this but I’m not going down without a fight,” Rogers allegedly texted. “These commies need to be told what’s up.”

FBI chief Christopher Wray issued a warning in early March that domestic terrorism is a growing threat in the United States, as exemplified by the attack on the Capitol on January 6th. In the wake of that attack, the FBI and other federal law enforcement agencies have prioritized prosecutions against domestic groups plotting violent attacks.

Rogers and Copeland were indicted on July 7th and face charges of conspiracy to destroy by fire or explosive a building used in or affecting interstate commerce. Rogers also faces charges for the illegal possession of firearms and explosives, and Copeland has been charged with destroying evidence.

“Firebombing your perceived political opponents is illegal and does not nurture the sort of open and vigorous debate that created and supports our constitutional democracy,” U.S. Attorney Stephanie Hinds said, according to CBS’ Bay Area affiliate, in the understatement of the year.

New Anthony Bourdain documentary deepfakes his voice

2021-07-15T15:01:22-04:00

In a new documentary, Roadrunner, about the life and tragic death of Anthony Bourdain, there are a few lines of dialogue in Bourdain’s voice that he might not have ever said out loud.

Filmmaker Morgan Neville used AI technology to digitally re-create Anthony Bourdain’s voice and have the software synthesize the audio of three quotes from the late chef and television host, Neville told the New Yorker.

The deepfaked voice was discovered when the New Yorker’s Helen Rosner asked how the filmmaker got a clip of Bourdain’s voice reading an email he had sent to a friend. Neville said he had contacted an AI company and supplied it with a dozen hours of Bourdain speaking.

“ … and my life is sort of shit now. You are successful, and I am successful, and I’m wondering: Are you happy?” Bourdain wrote in an email, and an AI algorithm later narrated an approximation of his voice.

You can hear the line in the documentary’s trailer linked below, right around the 1:30 mark. The algorithm’s generation of Bourdain’s voice is especially audible when it says, “and I am successful.”

Neville told Rosner that there were three lines of dialogue that he wanted Bourdain’s voice to orate, but he couldn’t find previous audio to string together or make it work otherwise.

There’s no shortage of companies that can achieve this kind of AI voice replication, and there’s actually a burgeoning industry of companies that can specifically generate voices for video game characters or allow you to clone your own voice.

But whether it’s ethical to clone a dead person’s voice and have them say things they hadn’t gotten on tape when they were alive is another question, and one Neville doesn’t seem too concerned with.

“We can have a documentary-ethics panel about it later,” he told the New Yorker.

Black teen barred from skating rink by inaccurate facial recognition

2021-07-15T14:37:02-04:00

A facial recognition algorithm used by a local roller skating rink in Detroit wouldn’t let teen Lamya Robinson onto the premises, and accused her of previously getting into a fight at the establishment.

But Robinson had never even been to the rink.

The facial recognition system had incorrectly matched her to another patron, she told Fox 2 Detroit. The rink removed her from the building and put her outside alone, her family says.

“To me, it’s basically racial profiling,” Juliea Robinson, her mother, told the TV station. “You’re just saying every young Black, brown girl with glasses fits the profile and that’s not right.”

The harms of facial recognition systems deployed in businesses and by police have been slowly coming to light as the technology is more widely used. Research into these algorithms has shown that they are far less accurate when distinguishing between the faces of Black people, women, and children, which might help explain the error faced by Lamya Robinson.

The highest-profile case of facial recognition leading to a wrongful arrest was also in Detroit, in the case of Robert Williams. Williams was arrested and detained for 30 hours in January 2020, after being accused of shoplifting from a Shinola watch store. He testified in front of the House Judiciary Committee, urging for legislators to adopt a moratorium on the technology introduced as legislation in June 2020.

“I don’t want anyone to walk away from my testimony thinking that if only the technology was made more accurate, its problems would be solved,” Williams said in his testimony. “Even if this technology does become accurate at the expense of people like me, I don’t want my daughters’ faces to be part of some government database.”

The disparity in racial and gender accuracy, as well as the invasive nature of the technology, has led to civil rights organizations and politicians calling for bans. The American Civil Liberties Union has called for nationwide bans and is suing the Detroit Police Department on behalf of Williams for its misuse of the technology. Some states like Maine have already begun to limit police use of the technology. However, only Portland, Oregon, currently has laws limiting how private businesses can use facial recognition.

Civil rights nonprofit Fight for the Future announced that more than 35 other organizations had joined it in demanding that retailers stop using facial recognition in their stores. The group reiterated its position today after the report of Lamya Robinson’s experience getting kicked out of the skating rink.

“This is exactly why we think facial recognition should be banned in public places,” wrote Fight for the Future’s director of campaign and operations Caitlin Seeley George in a press release. “It’s also not hard to imagine what could have happened if police were called to the scene and how they might have acted on this false information.”

The new Razer Blade 17 packs more GPU power into the same slim chassis

2021-07-14T11:00:00-04:00

Razer’s largest laptop is getting big upgrades. The new mid-2021 refresh of the Blade 17 will add newer CPUs, higher-powered GPUs, and an updated webcam.

The new Blade 17, which drops the “Pro” designation of earlier models, catches the larger laptop up to the Blade 15 Advanced released earlier this year. It now comes equipped with the same Intel 11th Gen H-Series processors, with a choice of the Core i7 11800H (2.3GHz base clock, 4.6GHz boost clock), or the top-tier Core i9 11900H (2.5GHz base, 4.9GHz boost), which isn’t available on Razer’s smaller models.

The laptop also features GPUs from Nvidia’s latest RTX lineup, including the 3060, 3070, and the 3080, with up to 16GB of video memory. Crucially, Razer has increased the Blade 17’s GPUs’ total graphics power to 130 watts, up from the 95 watts of the Blade 15 Advanced. That additional power means more heat generated, which might be mitigated in part by the Blade 17’s larger chassis. But we can’t say for sure whether heat will be an issue until a full test of the laptop.

Speaking of the laptop’s size, it stays the same with dimensions of 0.78 inches (19.99mm) x 15.5 inches (395mm) x 10.24 inches (260mm) and weighs 6.06 pounds.

The new Blade 17. | Image: Razer" data-portal-copyright="Image: Razer" />

Razer has built the Blade 17’s RAM and storage drives to be user-upgradeable. The laptop ships with a 1TB PCIe NVME drive, with an extra M.2 slot for expanded storage, and up to 32GB of RAM. The motherboard supports up to 32GB of RAM and 2TB of storage.

The Blade 17’s display options range from a 1440p display with a 165Hz refresh rate to a 1080p resolution with a quicker 360Hz refresh. One variant of the Blade 17 also packs a 4K touch display, though it only has a refresh rate of 120Hz. The sweet spot for most people is probably the 1440p model.

Some of the new features that round out the Blade 17 are the addition of anti-fingerprint coating previously added to the Blade 15 Advanced as well as an upgraded 1080p webcam with Windows Hello. The speaker system has also been upgraded, now with two speakers on each side of the laptop instead of just one.

The Blade 17 will charge through USB-C at 100W and will ship with Razer’s standard 230W charger.

There are also upgrades to the Blade 15 Base model as well, including the option of an i7-11800H processor and two PCIe 4.0 SSD slots.

You can preorder the Blade 17 and Blade 15 starting today, and Razer expects the laptops to ship in July–August.

Retail stores are packed with unchecked facial recognition, civil rights organizations say

2021-07-14T05:30:00-04:00

More than 35 organizations are demanding top US retailers cease using facial recognition to identify shoppers and employees in their stores, which companies have used to deter theft and identify shoplifters.

The campaign is aptly named Ban Facial Recognition in Stores, and has identified stores that have committed to not using facial recognition, like Walmart, Home Depot, and Target. It is now pressuring companies currently using the technology, or those who might use it in the future. Some companies that are currently using the technology, according to the website, include Apple, Lowe’s, Albertsons, Macy’s, and Ace Hardware. Apple tells The Verge that it does not use facial recognition in its stores.

Companies that might use it in the future include McDonalds, Walgreens, and 7-Eleven. A full list can be found on the website.

Privacy non-profit Fight for the Future organized the campaign last month as part of its larger Ban Facial Recognition project, and but has now garnered support from civil rights organizations like Mijente, Public Citizen, and Data for Black Lives, as well as others like the Consumer Federation of America and the Tor Project.

Tawana Petty, the national organizing director at Data for Black Lives, explains in the campaign’s press release that Detroit has implemented Project Green Light, which places surveillance cameras with facial recognition in more than 700 businesses.

“These cameras using facial recognition are monitored at real-time crime centers, police precincts, and on officers’ mobile devices 24/7,” she said. “It’s difficult to explain the psychological toll it takes on a community, knowing that your every move is being monitored by a racially biased algorithm with the power to yank your freedom away from you.”

Retailers have used invasive technologies like products that ping your phone’s Bluetooth and then catalog its unique MAC address, but AI-powered upgrades to video surveillance systems have allowed features like facial recognition to be more accessible than ever.

For instance, Rite-Aid quietly installed facial recognition cameras in hundreds of US stores, mostly in non-white and lower-income neighborhoods, Reuters reported in July 2020. The cameras scanned shoppers’ faces to try and find people in the store who were previously suspected of criminal activity and then sent alerts to security.

Walmart also reportedly used Clearview AI’s facial recognition more than 300 times, according to BuzzFeed News, and had previously used the technology to try and catch shoplifters. Now, it seems to be taking a different approach by using AI to find un-scanned goods in self-checkout kiosks.

Since these stores are private property, companies are not bound by state or local regulations banning government use of facial recognition which have been the most popular form of regulation. Only Portland, Oregon has included private businesses in a facial recognition ban.

Fight for the Future also points out that the coronavirus pandemic has expanded the use of facial recognition-adjacent surveillance in stores, which includes software for customer counting and social distancing.

Correction, July 14th, 2021, 2:35PM ET: This article previously stated that facial recognition company Kairos supported the Ban Facial Recognition in Stores campaign. However, the organization supporting the campaign is in fact Kairos Action, a civil rights nonprofit.

Update: July 14th, 2021, 10:15PM ET: Updated with statement from Apple.

Facebook groups can now designate Group Experts

2021-07-13T17:21:24-04:00

Facebook is introducing a new feature called “Group Experts” today, which will allow admins to designate which members are highly knowledgeable about the group’s topic.

Beyond the bragging rights, Group Experts will have a badge next to their name emblazoned with the proof of their elevated status. This badge will show up on posts, comments, and in the Q&A feature.

Group Experts will be selected by group administrators, and experts will only be given this extraordinary power once they have accepted the nomination. Admins can revoke the expert status, as well.

Facebook is testing ways to make it easier for admins to find these experts, by prompting specific users, currently in the gaming and fitness spaces, to add which games or activities they’re knowledgeable about. Group admins can search these topics to find their new Group Experts. This tactic works as a growth driver for groups as well, since the new experts will be able to invite followers who have recently interacted with their main page to join the group too.

However, if a group nominates an untrustworthy expert or if that group is willfully engaged in posting misinformation, it will exclusively be up to the admins to remove them from their expert status. For instance, groups that push anti-vax agendas would likely not nominate conventional doctors or public health officials as experts. So while the “expert” tag designates the trust of admins, it doesn’t objectively qualify someone as a real expert.

When asked about additional moderation tools to ensure Group Experts provide accurate information, a Facebook spokesperson listed the established ways the company deals with misinformation on the platform, like third-party fact checkers and reducing distribution for content flagged as misleading. They added that members can always report content to group admins.

Amazon reportedly doesn’t want employees to know when they’re on performance plans

2021-07-09T17:52:08-04:00

Managers at Amazon are instructed not to warn employees that they are being monitored under the company’s “Focus” performance management tool, according to documents obtained by The Seattle Times.

Focus is supposed to be a program where managers document how they are coaching employees and bringing their performance up to snuff. But in documents obtained by the Times, managers were instructed to not discuss Focus with employees, and instead they are supposed to more generally explain that the worker isn’t meeting expectations and how to improve. Managers would only be able to tell an employee they’re on the plan if they’re asked directly.

This means an employee could be on track to losing their job without ever knowing that their performance is under intense corporate scrutiny, and some who found out they were in the program said they have never gotten feedback on their deficiencies at all.

If Amazon employees don’t improve while unknowingly in the Focus program, they are then placed into the “Pivot” program, according to previous reporting from Business Insider. Employees told Business Insider that if they were placed in Pivot, they were either offered a severance package or given a chance to be put on a performance improvement plan.

Performance improvement plans, colloquially known as PIPs in the tech industry, are supposedly meant to bring underperforming employees back into the fold. But previous reporting by The Seattle Times also uncovered that Amazon expects a third of those placed on PIPs to leave the company, furthering its goal to filter out the lowest performing 6 percent of the company each year.

Amazon’s Focus tactic puts both employees and managers in a tough spot. It’s difficult for an employee to know the severity of their situation if they don’t know that they’re formally in the Focus program. And managers told the Times that they felt that keeping the program secret was counterproductive to employees getting on the right track.

“It’s a hidden way of weeding out people who are not part of the clique,” one Amazon employee told the Times.

US sanctions a Chinese facial recognition company with Silicon Valley funding

2021-07-09T13:42:59-04:00

The US Department of Commerce has sanctioned 14 Chinese tech companies over links to human rights abuses against Uyghur Muslims in Xinjiang, including one backed by a top Silicon Valley investment firm.

DeepGlint, also known as Beijing Geling Shentong Information Technology Co., Ltd., is a facial recognition company with deep ties to Chinese police surveillance, and funding from US-based Sequoia Capital. Today the Commerce Department added it to its Entity List, which restricts US companies from doing business with listed firms without a special license. Sequoia did not immediately respond to a request for comment.

DeepGlint co-founded a facial recognition lab in 2018 with Chinese authorities in Urumqi, the capital of Xinjiang, according to the South China Morning Post. It has also gained international bragging rights through the US National Institute of Standards and Technology’s (NIST) Face Recognition Vendor Test. DeepGlint claimed top accuracy in the test as of January 2021, giving it a potent marketing tool in the security and surveillance industry.

Bill Gates also referred to the startup as “very cool” in 2014

While DeepGlint has been accepted for a public offering on Shanghai’s STAR stock exchange, the firm hasn’t seen the commercial success of other AI startups in the country, explained Jeffrey Ding in his ChinAI newsletter last month. Since the firm is so heavily invested in government work, it has to follow slow government procurement cycles and is unlikely to score huge infrastructure projects, Ding writes.

Sequoia Capital has funded another company that later ended up on the Entity List. In 2020, Sequoia-backed Yitu Technology was added to the list for similar human rights abuses. Sequoia invested in DeepGlint back in 2014, before China’s genocide of Uyghurs had come to light. (The same year, Bill Gates also referred to the startup as “very cool,” according to KrAsia.)

The Commerce Department also sanctioned Xinjiang Lianhai Chuangzhi Company and Chengdu Xiwu Security System Alliance, two subsidiaries of Chinese military contractors. They both offer surveillance equipment and services, according to their websites and academic reports. Xinjiang Lianhai Chuangzhi Company created an AI-powered checkpoint system that is able to track Uyghurs as they move around cities, according to a report from the Italian Institute for International Political Studies.

Another firm sanctioned today is Leon Technology, a surveillance company that was controlled by Chinese AI giant SenseTime until its role providing oppressive technology in Xinjiang was reported in 2019. SenseTime then divested its 51 percent stake in it.

The Commerce Department sanctions also included nine other Chinese companies for national security reasons, as well as companies in Iran, Russia, and Canada, among other countries.

Minimum broadband speeds are likely too low, government watchdog says

2021-07-08T14:07:07-04:00

Since 2015, the Federal Communications Commission has maintained that minimum broadband speed should stay the same: 25 megabits per second for downloading files, and 3 megabits per second for uploading.

Now, a government watchdog is telling the FCC to re-evaluate these rules in a new report, saying small businesses in 2021 require more bandwidth to effectively run their operations. The report, produced by the Government Accountability Office, also shows examples of small businesses around the country which aren’t even getting the minimum speed required by the FCC.

One Vermont inn pays $78 per month for 10 Mbps speeds

This isn’t the first time the FCC has faced heat over its broadband minimums. Earlier this year a group of senators called for the agency to raise the bottom limit to 100 Mbps for both download and upload speed.

The FCC has basically admitted that it has very little reliable data on whether people are getting the mandated minimum speeds, or access at all. Its broadband coverage maps are based on self-reported data from internet providers, who are incentivized to overstate their coverage to shirk regulation. It has attempted to create speed test apps to get more direct information, and collected complaints directly from consumers.

Anonymized data visualized by The Verge paints a much bleaker picture than the FCC has admitted, with many counties reporting less than 10 percent broadband availability at minimum speeds.

But the GAO contends that there’s enough evidence to reconsider these rules. Reports from small businesses show that many want a minimum of 100 Mbps download speed, and up to 1 gigabyte per second for offices with dozens of employees. Even Zoom recommends at least 3.8 Mbps of upload speed to make high-definition video calls, which exceeds the federal minimum.

Business owners also told the GAO that while they might be able to access broadband, it’s expensive and unreliable. One Vermont inn and spa owner said that they pay $78 per month for substandard 10 Mbps speeds, and upgrading to 40 Mbps would cost a prohibitive $335 per month. Others said they had to rely on expensive satellite internet services to have any reliable access to the internet.

Raising this minimum benchmark to 100 Mbps down / 10 Mbps up would have long-term effects for people in rural communities who suffer most from these internet deserts, the GAO writes. The change would lower the percentage of rural Americans who are considered to have acceptable broadband coverage from 83 percent to 67 percent, a strong argument that providers need to invest more in rural infrastructure.