Archives April 2019

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.

Facebook CEO Mark Zuckerberg once considered making deals with third-party developers just to help him find out how much users’ data is worth, NBC News reported on Tuesday. The report, which cites 4,000 leaked pages of internal documents, shines a light on the way senior company executives viewed attaching a dollar sign to sensitive user data, despite Facebook’s public commitment to protect such information. From the report:

In the same week, Zuckerberg floated the idea of pursuing 100 deals with developers “as a path to figuring out the real market value” of Facebook user data and then “setting a public rate” for developers. “The goal here wouldn’t be the deals themselves, but that through the process of negotiating with them we’d learn what developers would actually pay (which might be different from what they’d say if we just asked them about the value), and then we’d be better informed on our path to set a public rate,” Zuckerberg wrote in a chat. Facebook told NBC News that it was exploring ways to build a sustainable business, but ultimately decided not to go forward with these plans.

Zuckerberg was unfazed by the potential privacy risks associated with Facebook’s data-sharing arrangements. “I’m generally skeptical that there is as much data leak strategic risk as you think,” he wrote in the email to Lessin. “I think we leak info to developers but I just can’t think of any instances where that data has leaked from developer to developer and caused a real issue for us.”

The report also outlines how PR people at Facebook tries to spin things. An excerpt: In a March 2014 email discussing Zuckerberg’s keynote speech at the event, where he was due to announce the removal of developers’ access to friends’ data, Jonny Thaw, a director of communications, wrote that it “may be a tough message for some developers as it may inhibit their growth.” “So one idea that came up today was potentially talking in the keynote about some of the trust changes we’re making on Facebook itself. So the message would be: ‘trust is really important to us — on Facebook, we’re doing A, B and C to help people control and understand what they’re sharing — and with platform apps we’re doing D, E and F.'” If that doesn’t work, he added, “we could announce some of Facebook’s trust initiatives in the run up to F8” to make the changes for developers “seem more natural.”

Despite Facebook’s repeated warnings that law enforcement is required to use “authentic identities” on the social media platform, cops continue to create fake and impersonator accounts to secretly spy on users. By pretending to be someone else, cops are able to sneak past the privacy walls users put up and bypass legal requirements that might require a warrant to obtain that same information.

EFF is now calling on Facebook to escalate the matter with law enforcement in the United States. Facebook should take the following actions to address the proliferation of fake/impersonator Facebook accounts operated by law enforcement, in addition to suspending the fake accounts. As part of its regular transparency reports, Facebook should publish data on the number of fake/impersonator law enforcement accounts identified, what agencies they belonged to, and what action was taken. When a fake/impersonator account is identified, Facebook should alert the users and groups that interacted with the account whether directly or indirectly.

The article also suggests updating Facebook’s Terms of Service to explicitly prohibit fake/impersonator profiles by law enforcement groups, and updating Facebook pages of law enforcement groups to inform visitors when those groups have a written policy allowing fake/impersonator law enforcement accounts. “These four changes are relatively light lifts that would enhance transparency and establish real consequences for agencies that deliberately violate the rules…”

“Facebook’s practice of taking down these individual accounts when they learn about them from the press (or from EFF) is insufficient to deter what we believe is a much larger iceberg beneath the surface.”

After crunching data on 4.5 billion miles of driving, road-safety analytics company Zendrive concludes there’s a new threat which just last year claimed the lives of 6,227 pedestrians: drivers “under the influence of a smartphone.”

The study points out that drunk driving fatalities peak after midnight, while distracted driving happens all day, conluding that distracted driving is now a bigger threat than drunk driving.

“Phone addicts are the new drunk drivers,” Zendrive concludes bluntly in its annual distracted driving study. The big picture: The continued increase in unsafe driving comes despite stricter laws in many states, as well as years of massive ad campaigns from groups ranging from cell phone carriers to orthopedic surgeons. “They hide in plain sight, blatantly staring at their phones while driving down the road,” Zendrive says in the study.

And it’s a growing problem. Over just the past year, Zendrive, which analyzes driver behavior for fleets and insurers, said the number of hardcore phone addicts doubled, now accounting for one in 12 drivers. If the current trend continues, that number will be one in five by 2022.

The report concludes drivers are 10 percent more distracted this year than last — and that phone addicts have their eyes off the road for 28% of their drive. Yet when asked to describe their driving, 93% of phone addicts said they believed they were “safe” — or “extremely safe” — drivers.

One even insisted that they never texted while driving, “but I like to FaceTime my friends while driving since it makes time go by faster.”

In the past couple of years, genetic-testing companies like Ancestry and 23andMe have become popular for finding out family history and DNA information. More than 12 million Americans have sent in their DNA to be analyzed to companies like 23andMe and AncestryDNA. The spit-in-tube DNA you send in is anonymized and used for genetic drug research and both sites have been selling the data to third-party companies, like P&G Beauty and Pepto-Bismol, and universities, like The University of Chicago, for some time. In fact, just last week major pharmaceutical giant, GlaxoSmithKline, announced a $300 million deal with 23andMe. The deal entails that they can use the data to analyze the stored sample, investigate new drugs to develop and genetic data for how patients are selected for clinical trials. Both 23andMe and Ancestry said that they will not share genetic information freely, without a court order, but people are welcome to share the information online themselves sometimes in order to find lost relatives or biological parents.

On Wednesday, the Illinois State Senate passed the Keep Internet Devices Safe Act, a bill that would ban manufacturers of devices that can record audio from doing so remotely without disclosing it to the customer. But after lobbying from trade associations that represent the interests of Google, Amazon — makers of the microphone-enabled Google Home and Alexa smart speakers, respectively — and Microsoft, among other companies, the interests of big tech won out… In its current, neutered form, the bill provides exclusive authority to the Attorney General to enforce the Act, which means regular citizens won’t be able to bring forward a case regarding tech giants recording them in their homes.

Ars Technica notes the move comes after Amazon admitted thousands of their employees listen to Alexa recordings — “something not mentioned in Echo’s terms of service or FAQ pages.”

Vice points out that sometimes those recordings are shared “even after users opt out of having their data used in the program.”

Police have used information from the search giant’s Sensorvault database to aid in criminal cases across the country, according to a report Saturday by The New York Times. The database has detailed location records from hundreds of millions of phones around the world, the report said. It’s meant to collect information on the users of Google’s products so the company can better target them with ads, and see how effective those ads are. But police have been tapping into the database to help find missing pieces in investigations.

Law enforcement can get “geofence” warrants seeking location data. Those kinds of requests have spiked in the last six months, and the company has received as many as 180 requests in one week, according to the report…. For geofence warrants, police carve out a specific area and time period, and Google can gather information from Sensorvault about the devices that were present during that window, according to the report. The information is anonymous, but police can analyze it and narrow it down to a few devices they think might be relevant to the investigation. Then Google reveals those users’ names and other data, according to the Times…

[T]he AP reported last year that Google tracked people’s location even after they’d turned off location-sharing on their phones.

Google’s data dates back “nearly a decade,” the Times reports — though in a statement, Google’s director of law enforcement and information security insisted “We vigorously protect the privacy of our users while supporting the important work of law enforcement.” (The Times also interviewed a man who was arrested and jailed for a week last year based partly on Google’s data — before eventually being released after the police found a more likely suspect.)

After April 15, inmates at the Adult Detention Center in Lowndes County, Mississippi will no longer be allowed to visit with family members face to face. Newton County, Missouri, implemented an in-person visitor ban last month. The Allen County Jail in Indiana phased out in-person visits earlier this year. All three changes are part of a nationwide trend toward “video visitation” services. Instead of seeing their loved ones face to face, inmates are increasingly limited to talking to them through video terminals. Most jails give family members a choice between using video terminals at the jail — which are free — or paying fees to make calls from home using a PC or mobile device.

Even some advocates of the change admit that it has downsides for inmates and their families. Ryan Rickert, jail administrator at the Lowndes County Adult Detention Center, acknowledged to The Commercial Dispatch that inmates were disappointed they wouldn’t get to see family members anymore. Advocates of this approach point to an upside for families: they can now make video calls to loved ones from home instead of having to physically travel to the jail. These services are ludicrously expensive. Video calls cost 40 cents per minute in Newton County, 50 cents per minute in Lowndes County, and $10 per call in Allen County. Outside of prison, of course, video calls on Skype or FaceTime are free.

These “visitation” services are often “grainy and jerky, periodically freezing up altogether,” reports Ars. As for why so many jails are adopting them, it has a lot to do with money. “In-person visits are labor intensive. Prison guards need to escort inmates to and from visitation rooms, supervise the visits, and in some cases pat down visitors for contraband. In contrast, video terminals can be installed inside each cell block, minimizing the need to move inmates around the jail.” The video-visitation systems also directly generate revenue for jails.

Last summer, an AP investigation found that Google’s location tracking remains on even if you turn it off in Google Maps, Search, and other apps. Research from Vanderbilt professor Douglas Schmidt found that Google engages in “passive” data collection, often without the user’s knowledge. His research also showed that Google utilizes data collected from other sources to de-anonymize existing user data.

Digital Content Next, the trade association of online publishers, surveyed a nationally representative sample to find out what people expect from Google — and, as with a similar study we conducted last year about Facebook, the results were unsettling.

Our findings show that many of Google’s data practices deviate from consumer expectations. We find it even more significant that consumer’s expectations are at an all-time low even after 2018, a year in which awareness around consumer privacy reached peak heights.

The results of the study are consistent with our Facebook study: People don’t want surveillance advertising. A majority of consumers indicated they don’t expect to be tracked across Google’s services, let alone be tracked across the web in order to make ads more targeted.

There was only one question where a small majority of respondents felt that Google was acting according to their expectations. That was about Google merging data from search queries with other data it collects on its own services. They also don’t expect Google to connect the data back to the user’s personal account, but only by a small majority. Google began doing both of these in 2016 after previously promising it wouldn’t.