Researchers from Northeastern Unviersity, the University of Southern Carolina, and tech accountability non-profit Upturn have released a paper that says Facebook's ad delivery system itself can steer ads intended to be inclusive toward discrimination without explicit intent. "In a paper titled, 'Discrimination through optimization: How Facebook's ad delivery can lead to skewed outcomes,' co-authors Muhammad Ali, Piotr Sapiezynski, Miranda Bogen, Aleksandra Korolova, Alan Mislove, and Aaron Rieke find that advertiser budgets and ad content affect ad delivery, skewing it along gender and racial lines even when neutral ad targeting settings are used," reports The Register. From the report: The researchers found that Facebook ads tend to be shown to men because women tend to click on ads more often, making them more expensive to reach through Facebook's system. That divide becomes apparent when ad budgets are compared, because the ad budget affects ad distribution. As the paper explains, "the higher the daily budget, the smaller the fraction of men in the audience." Such segregation may be appropriate and desirable for certain types of marketing pitches, but when applied to credit, employment and housing ads, the consequences can be problematic.

Ad content -- text and images -- also has a strong effect on whether ads get shown to men or women, even when the bidding strategy is the same and gender-agnostic targeting is used. In particular, the researchers found images had a surprisingly large effect on ad delivery. Ad URL destination has some effect -- an ad pointing to a bodybuilding site and an ad pointing to a cosmetics site had a baseline delivery distribution of 48 percent men and 40 percent men respectively. The addition of a title and headline doesn't change that much. But once the researchers added an image to the ad, the distribution pattern changed, with the bodybuilding site ad reaching an audience that was 75 percent male and the cosmetics ad reaching an audience that was 90 percent female. According to the researchers, their tests suggest, "Facebook has an automated image classification mechanism in place that is used to steer different ads towards different subsets of the user population." "In terms of credit, employment and housing ads, the problem with this system is that it discriminates where it shouldn't: Five ads for lumber industry jobs were delivered to an audience that was more than 90 percent men and more than 70 percent white; five ads for janitorial work were delivered to an audience that was more than 65 percent women and 75 percent black," the report adds. "Housing ads also showed a racial skew."

The latest findings come after years of criticism of Facebook's ad system. Last month, Facebook announced changes to the platform intended to prevent advertisers from deploying unfair credit, employment and housing ads. One week later, the U.S. Department of Housing and Urban Development sued Facebook for violating the Fair Housing Act.

An anonymous reader quotes a report from Ars Technica: Democrats in the U.S. House of Representatives yesterday rejected Republican attempts to weaken a bill that would restore net neutrality rules. The House Commerce Committee yesterday approved the "Save the Internet Act" in a 30-22 party-line vote, potentially setting up a vote of the full House next week. The bill is short and simple -- it would fully reinstate the rules implemented by the Federal Communications Commission under then-Chairman Tom Wheeler in 2015, reversing the repeal led by FCC Chairman Ajit Pai in 2017.

Commerce Committee Republicans repeatedly introduced amendments that would weaken the bill but were consistently rebuffed by the committee's Democratic majority. "The Democrats beat back more than a dozen attempts from Republicans to gut the bill with amendments throughout the bill's markup that lasted 9.5 hours," The Hill reported yesterday. Republican amendments would have weakened the bill by doing the following: Exempt all 5G wireless services from net neutrality rules; Exempt all multi-gigabit broadband services from net neutrality rules; Exempt from net neutrality rules any ISP that builds broadband service in any part of the U.S. that doesn't yet have download speeds of at least 25Mbps and upload speeds of at least 3Mbps; Exempt from net neutrality rules any ISP that gets universal service funding from the FCC's Rural Health Care Program; Exempt ISPs that serve 250,000 or fewer subscribers from certain transparency rules that require public disclosure of network management practices; and Prevent the FCC from limiting the types of zero-rating (i.e., data cap exemptions) that ISPs can deploy. An additional Republican amendment "would have imposed net neutrality rules but declared that broadband is an information service, [preventing] the FCC from imposing any other type of common-carrier regulations on ISPs," reports Ars Technica. "The committee did approve a Democratic amendment to exempt ISPs with 100,000 or fewer subscribers from the transparency rules, but only for one year."

U.S. authorities gathered information about Huawei through secret surveillance that they plan to use in a case accusing the Chinese telecom equipment maker of sanctions-busting and bank fraud, prosecutors said on Thursday. From a report: Assistant U.S. Attorney Alex Solomon said at a hearing in federal court in Brooklyn that the evidence, obtained under the U.S. Foreign Intelligence Surveillance Act (FISA), would require classified handling. The government notified Huawei in a court filing on Thursday of its intent to use the information, saying it was "obtained or derived from electronic surveillance and physical search," but gave no details. The United States has been pressuring other countries to drop Huawei from their cellular networks, worried its equipment could be used by Beijing for spying. The company says the concerns are unfounded. Brian Frey, a former federal prosecutor who is not involved in the Huawei case, said FISA surveillance, which requires a warrant from a special court, is generally sought in connection with suspected espionage.

Microsoft has announced that starting with the Windows 10 May 2019 Update, which will hit general availability late next month, users will no longer be forced to install new Windows 10 feature updates as they become available. From a report: This comes after feedback from users who have had countless issues with updates breaking programs, losing files, and installing at inconvenient times. Microsoft has been working hard to improve Windows Update, and while the system is better than it was at launch in 2015, it's still not perfect. Now, users will have the option to not have to deal with feature updates when they are released.

What Microsoft is doing here is splitting Windows Update in two. The normal "check for updates" button will now only function for security and monthly patches. Feature updates now get their own area in Windows Update where the user can initiate the download and install process for the latest feature update available. If the user doesn't want to initiate that process, they don't have to. The user will be alerted that a new feature update is available every now and then, but at no point will the user be forced to install that update, as long as the version of Windows 10 they're currently running is still in support.

Steven J. Vaughan-Nichols, writing for ComputerWorld : Of course, at one time, to get any work done with a computer, you first had to learn a lot, about computers, operating systems, commands and more. Eventually, "friendly" became the most important adverb in computing circles, and we've reached the point in user-friendliness that people don't even talk about it anymore. Today, Google has shown with its Chrome OS that most of us can pretty much do anything we need to do on a computer with just a web browser. But Google's path is not Microsoft's path. Instead, it's moving us first to Windows as desktop as a service (DaaS) via Microsoft Managed Desktop (MMD). This bundles Windows 10 Enterprise, Office 365 and Enterprise Mobility + Security and cloud-based system management into Microsoft 365 Enterprise.

The next step, Windows Virtual Desktop, enables companies to virtualize Windows 7 and 10, Office 365 ProPlus apps and other third-party applications on Azure-based virtual machines. If all goes well, you'll be able to subscribe to Windows Virtual Desktop this fall. Of course, Virtual Desktop is a play for business users -- for now. I expect Virtual Desktop to be offered to consumers in 2020. By 2025, Windows as an actual desktop operating system will be a niche product. Sound crazy? Uh, you do know that Microsoft already really, really wants you to "rent" Office 365 rather than buy Office 2019, don't you?

But what about games, you say? We'll always have Windows for games! Will we? Google, with its Google Stadia gaming cloud service, is betting we're ready to move our games to the cloud as well. It's no pipe dream. Valve has been doing pretty well for years now with its Steam variation on this theme. So where is all this taking us? I see a world where the PC desktop disappears for all but a few. Most of us will be writing our documents, filling out our spreadsheets and doing whatever else we now do on our PCs via cloud-based applications on smart terminals running Chrome OS or Windows Lite. If you want a "real" PC, your choices are going to be Linux or macOS.

The pilots of a doomed Ethiopian Airlines jet followed all of Boeing's recommended procedures when the plane started to nose dive but still couldn't save it, according to findings from a preliminary report released Thursday by the Ethiopian government. From a report: The plane crashed just six minutes after taking off from Addis Ababa, killing all 157 people on board. The report, based on flight data and cockpit voice recorders on the Boeing 737 Max 8, was not released in full. Boeing declined to comment pending its review of the report on the March 10 crash. The Max 8 has been under scrutiny since a Lion Air flight crashed off the coast of Indonesia under similar circumstances in October. Thursday's revelations raise questions about repeated assertions by Boeing and U.S. regulators that pilots could regain control in some emergencies by following steps that include turning off an anti-stall system designed specifically for the Max, known by its acronym, MCAS. Investigators are looking into the role of MCAS, whose functions include automatically lowering the plane's nose to prevent an aerodynamic stall. The Max has been grounded worldwide pending a software fix that Boeing is rolling out, which still needs to be approved by the Federal Aviation Administration and other regulators. Further reading: Flawed Analysis, Failed Oversight: How Boeing, FAA Certified the Suspect 737 MAX Flight Control System.

Lots of things can hurt your social credit in China. Failing to repay your debts, plagiarizing academic articles and building a debt-laden tech empire and then fleeing to another country, to name a few random examples. One province now wants to add another "discredited behavior" that seems much more harmless: Switching jobs too often. From a report: Zhejiang is pushing to build a local social credit system that will, among other things, deem residents a "discredited" person if they move from job to job too frequently, according to a local TV report. "If someone keeps quitting and landing new jobs, his social credit will definitely be a problem," Zhejiang official Ge Pingan said at a local forum, addressing a complaint from one company's human resources department about being unable to do anything when employees want to leave. Ge didn't specify how "frequently" is too frequent, but he said the upcoming system will put restrictions on both companies and individual workers.

Amazon is joining the race to provide broadband internet access around the globe via thousands of satellites in low Earth orbit, newly uncovered filings show. From a report: The effort, code-named Project Kuiper, follows up on last September's mysterious reports that Amazon was planning a "big, audacious space project" involving satellites and space-based systems. The Seattle-based company is likely to spend billions of dollars on the project, and could conceivably reap billions of dollars in revenue once the satellites go into commercial service. It'll take years to bring the big, audacious project to fruition, however, and Amazon could face fierce competition from SpaceX, OneWeb and other high-profile players.

Project Kuiper's first public step took the form of three sets of filings made with the International Telecommunications Union last month by the Federal Communications Commission on behalf of Washington, D.C.-based Kuiper Systems LLC. The ITU oversees global telecom satellite operations and eventually will have to sign off on Kuiper's constellation. The filings lay out a plan to put 3,236 satellites in low Earth orbit -- including 784 satellites at an altitude of 367 miles (590 kilometers); 1,296 satellites at a height of 379 miles (610 kilometers); and 1,156 satellites in 391-mile (630-kilometer) orbits. In response to GeekWire's inquiries, Amazon confirmed that Kuiper Systems is actually one of its projects.

An anonymous reader quotes a report from The New York Times: Australia passed sweeping legislation Thursday that threatens huge fines for social media companies and jail for their executives (Warning: source may be paywalled; alternative source) if they fail to rapidly remove "abhorrent violent material" from their platforms. The law -- strongly opposed by the tech industry -- puts Australia at the forefront of a global movement to hold companies like Facebook and YouTube accountable for the content they host. It comes less than a month after a gunman, believed to be an Australian white nationalist, distributed a hate-filled manifesto online before using Facebook to live-stream the massacre of 50 people at two mosques in Christchurch, New Zealand. Written quickly and without much input from technology companies or experts, the measure goes as far as any other democracy's attempt to punish multinational tech platforms for the behavior of their users. "The legislation criminalizes 'abhorrent violent material,' which it defines as videos that show terrorist attacks, murders, rape or kidnapping. Social media companies that fail to remove such content 'expeditiously' could face fines of up to 10 percent of their annual profit, and employees could be sentenced to up to three years in prison," the report adds. "Companies must also inform the police when illegal material is found."

"This law, which was conceived and passed in five days without any meaningful consultation, does nothing to address hate speech, which was the fundamental motivation for the tragic Christchurch terrorist attacks," said Sunita Bose, the managing director of the Digital Industry Group, an advocacy group representing Facebook, Google and other companies. "With the vast volumes of content uploaded to the internet every second, this is a highly complex problem that requires discussion with the technology industry, legal experts, the media and civil society to get the solution right -- that didn't happen this week."