At a moment when education technology firms are stockpiling sensitive information on millions of school children, safeguards for student data have broken down.
Lina Khan may set off a shift in how Washington regulates competition by filing cases in tech areas before they mature. She faces an uphill climb.
The move is a potential blow to Meta’s metaverse efforts and signals a shift in how the Federal Trade Commission is approaching tech deals.
The most urgent step the federal government can take is to ensure access to medication abortion.
In an interview, the chair of the Federal Trade Commission laid out some of her plans now that she has a Democratic majority at the agency.
Senior officials announced minor steps the administration would take to try to increase the supply of formula, even as they conceded that Americans might not see quick relief.
The confirmation of a third Democrat creates an opportunity for Lina Khan, the Federal Trade Commission’s chair, to advance efforts to rein in corporate power.
The N.F.L. franchise sent 102 pages of documents rebutting claims made by a former employee that the team failed to properly report ticket revenue and withheld refunds from fans.
In a 20-page letter sent to the F.T.C., a congressional committee passed along allegations from a former employee who said the team withheld ticket revenue from fans and the league.
The company’s first consumer protection lawsuit, filed Monday, claims a Cameroon man tricked would-be buyers using Gmail and other services.
The wireless carrier said that it was working with the F.B.I. and the Secret Service to investigate a recent wave of fraudulent messages, but said the source did not appear to be Russian hackers.
A California tech company that tried to solve the problem of malfunctioning ice cream machines claims in the suit that it was thwarted by McDonald’s.
Kurbo by WW, a weight loss app geared toward children, illegally collected data from users as young as 8 without their parents’ consent, the Federal Trade Commission said in a complaint.
The agency’s expected lawsuit against the deal for a health technology company would be the latest move by the Biden administration to quash corporate consolidation.
The deal, which was initially valued at $40 billion, encountered regulatory scrutiny, including an F.T.C. lawsuit.
The government can proceed with its claims that the company abused its monopoly power through acquisitions.
“You’re feeding America and going broke doing it”: After years of consolidation, four companies dominate the meatpacking industry, while many ranchers are barely hanging on.
A wide-ranging presidential order helped block a railroad merger and tackle supply-chain problems, and it is planting the seeds for bigger actions.
A wide-ranging presidential order helped block a railroad merger and tackle supply-chain problems, and it is planting the seeds for bigger actions.
The proposed deal would give Nvidia control over computing technology and designs that rival firms rely on.
President Biden asked the Federal Trade Commission to look into whether big oil companies are fueling a spike in gas prices.
President Biden asked the Federal Trade Commission to look into whether big oil companies are fueling a spike in gas prices.
Updating our war against the scammers and assorted interrupters.
Andy Parker, the father of a journalist killed in 2015, filed a complaint with the F.T.C. urging it to increase regulation on the tech giant, saying it failed to remove videos of his daughter’s killing.
Will the industry finally have its feet held to the fire?
If what faces Big Tech is anything like what happened to Big Tobacco, the road ahead is likely to be a yearslong battle over proposed rules and regulations.
How Jonathan Kanter, the Biden administration’s choice to be the Justice Department’s antitrust chief, became a progressive foe of Big Tech.
The incoming director of the Consumer Financial Protection Bureau, Rohit Chopra, is expected to use his position to rein in financial firms.
President Biden made his latest nomination to the Federal Trade Commission this week, tapping digital privacy expert Alvaro Bedoya to join the agency as it takes a hard look at the tech industry.
Bedoya is the founding director of the Center on Privacy & Technology at Georgetown’s law school and previously served as chief counsel for former Senator Al Franken and the Senate Judiciary Subcommittee on Privacy, Technology, and the Law. Bedoya has worked on legislation addressing some of the most pressing privacy issues in tech, including stalkerware and facial recognition systems.
In 2016, Bedoya co-authored a report titled “The Perpetual Line-Up: Unregulated Police Face Recognition in America,” a year-long investigation that dove deeply into the police use of facial recognition systems in the U.S. The 2016 report examined law enforcement’s reliance on facial recognition systems and biometric databases on a state level. It argued that regulations are desperately needed to curtail potential abuses and algorithmic failures before the technology inevitably becomes even more commonplace.
Bedoya also isn’t shy about calling out Big Tech. In a New York Times op-ed a few years ago, he took aim at Silicon Valley companies giving user privacy lip service in public while quietly funneling millions toward lobbyists to undermine consumer privacy. The new FTC nominee singled out Facebook specifically, pointing to the company’s efforts to undermine the Illinois Biometric Information Privacy Act, a state law that serves as one of the only meaningful checks on invasive privacy practices in the U.S.
Bedoya argued that the tech industry would have an easier time shaping a single, sweeping piece of privacy regulations with its lobbying efforts rather than a flurry of targeted, smaller bills. Antitrust advocates in Congress taking aim at tech today seem to have learned that same lesson as well.
“We cannot underestimate the tech sector’s power in Congress and in state legislatures,” Bedoya wrote. “If the United States tries to pass broad rules for personal data, that effort may well be co-opted by Silicon Valley, and we’ll miss our best shot at meaningful privacy protections.”
If confirmed, Bedoya would join big tech critic Lina Khan, a recent Biden FTC nominee who now chairs the agency. Khan’s focus on antitrust and Amazon in particular would dovetail with Bedoya’s focus on adjacent privacy concerns, making the pair a formidable regulatory presence as the Biden administration seeks to rein in some of the tech industry’s most damaging excesses.
What we leave behind.
Does this sound familiar? An app goes viral on social media, often including TikTok, then immediately climbs to the top of the App Store where it gains even more new installs thanks to the heightened exposure. That’s what happened with the recent No. 1 on the U.S. App Store, Fontmaker, a subscription-based fonts app which appeared to benefit from word-of-mouth growth thanks to TikTok videos and other social posts. But what we’re actually seeing here is a new form of App Store marketing — and one which now involves one of the oldest players in the space: Vungle.
Fontmaker, at first glance, seems to be just another indie app that hit it big.
The app, published by an entity called Mango Labs, promises users a way to create fonts using their own handwriting which they can then access from a custom keyboard for a fairly steep price of $4.99 per week. The app first launched on July 26. Nearly a month later, it was the No. 2 app on the U.S. App Store, according to Sensor Tower data. By August 26, it climbed up one more position to reach No. 1. before slowly dropping down in the top overall free app rankings in the days that followed.
By Aug. 27, it was No. 15, before briefly surging again to No. 4 the following day, then declining once more. Today, the app is No. 54 overall and No. 4 in the competitive Photo & Video category — still, a solid position for a brand-new and somewhat niche product targeting mainly younger users. To date, it’s generated $68,000 in revenue, Sensor Tower reports.
But Fontmaker may not be a true organic success story, despite its Top Charts success driven by a boost in downloads coming from real users, not bots. Instead, it’s an example of how mobile marketers have figured out how to tap into the influencer community to drive app installs. It’s also an example of how it’s hard to differentiate between apps driven by influencer marketing and those that hit the top of the App Store because of true demand — like walkie-talkie app Zello, whose recent trip to No. 1 can be attributed to Hurricane Ida
As it turns out, Fontmaker is not your typical “indie app.” In fact, it’s unclear who’s really behind it. Its publisher, Mango Labs, LLC, is actually an iTunes developer account owned by the mobile growth company JetFuel, which was recently acquired by the mobile ad and monetization firm Vungle — a longtime and sometimes controversial player in this space, itself acquired by Blackstone in 2019.
Through The Plug, mobile app developers and advertisers can connect to JetFuel’s network of over 15,000 verified influencers who have a combined 4 billion Instagram followers, 1.5 billion TikTok followers, and 100 million daily Snapchat views.
While marketers could use the built-in advertising tools on each of these networks to try to reach their target audience, JetFuel’s technology allows marketers to quickly scale their campaigns to reach high-value users in the Gen Z demographic, the company claims. This system can be less labor-intensive than traditional influencer marketing, in some cases. Advertisers pay on a cost-per-action (CPA) basis for app installs. Meanwhile, all influencers have to do is scroll through The Plug to find an app to promote, then post it to their social accounts to start making money.
So while yes, a lot of influencers may have made TikTok videos about Fontmaker, which prompted consumers to download the app, the influencers were paid to do so. (And often, from what we saw browsing the Fontmaker hashtag, without disclosing that financial relationship in any way — an increasingly common problem on TikTok, and area of concern for the FTC.)
Where things get tricky is in trying to sort out Mango Labs’ relationship with JetFuel/Vungle. As a consumer browsing the App Store, it looks like Mango Labs makes a lot of fun consumer apps of which Fontmaker is simply the latest.
JetFuel’s website helps to promote this image, too.
It had showcased its influencer marketing system using a case study from an “indie developer” called Mango Labs and one of its earlier apps, Caption Pro. Caption Pro launched in Jan. 2018. (App Annie data indicates it was removed from the App Store on Aug. 31, 2021…yes, yesterday).
Vungle, however, told TechCrunch “The Caption Pro app no longer exists and has not been live on the App Store or Google Play for a long time.” (We can’t find an App Annie record of the app on Google Play).
They also told us that “Caption Pro was developed by Mango Labs before the entity became JetFuel,” and that the case study was used to highlight JetFuel’s advertising capabilities. (But without clearly disclosing their connection.)
“Prior to JetFuel becoming the influencer marketing platform that it is today, the company developed apps for the App Store. After the company pivoted to become a marketing platform, in February 2018, it stopped creating apps but continued to use the Mango Labs account on occasion to publish apps that it had third-party monetization partnerships with,” the Vungle spokesperson explained.
In other words, the claim being made here is that while Mango Labs, originally, were the same folks who have long since pivoted to become JetFuel, and the makers of Caption Pro, all the newer apps published under “Mango Labs, LLC” were not created by JetFuel’s team itself.
“Any apps that appear under the Mango Labs LLC name on the App Store or Google Play were in fact developed by other companies, and Mango Labs has only acted as a publisher,” the spokesperson said.
There are reasons why this statement doesn’t quite sit right — and not only because JetFuel’s partners seem happy to hide themselves behind Mango Labs’ name, nor because Mango Labs was a project from the JetFuel team in the past. It’s also odd that Mango Labs and another entity, Takeoff Labs, claim the same set of apps. And like Mango Labs, Takeoff Labs is associated with JetFuel too.
Breaking this down, as of the time of writing, Mango Labs has published several consumer apps on both the App Store and Google Play.
On iOS, this includes the recent No. 1 app Fontmaker, as well as FontKey, Color Meme, Litstick, Vibe, Celebs, FITme Fitness, CopyPaste, and Part 2. On Google Play, it has two more: Stickered and Mango.
Most of Mango Labs’ App Store listings point to JetFuel’s website as the app’s “developer website,” which would be in line with what Vungle says about JetFuel acting as the apps’ publisher.
What’s odd, however, is that the Mango Labs’ app Part2, links to Takeoff Labs’ website from its App Store listing.
The Vungle spokesperson initially told us that Takeoff Labs is “an independent app developer.”
And yet, the Takeoff Labs’ website shows a team which consists of JetFuel’s leadership, including JetFuel co-founder and CEO Tim Lenardo and JetFuel co-founder and CRO JJ Maxwell. Takeoff Labs’ LLC application was also signed by Lenardo.
Meanwhile, Takeoff Labs’ co-founder and CEO Rhai Goburdhun, per his LinkedIn and the Takeoff Labs website, still works there. Asked about this connection, Vungle told us they did not realize the website had not been updated, and neither JetFuel nor Vungle have an ownership stake in Takeoff Labs with this acquisition.
Takeoff Labs’ website also shows off its “portfolio” of apps, which includes Celeb, Litstick, and FontKey — three apps that are published by Mango Labs on the App Store.
On Google Play, Takeoff Labs is the developer credited with Celebs, as well as two other apps, Vibe and Teal, a neobank. But on the App Store, Vibe is published by Mango Labs.
(Not to complicate things further, but there’s also an entity called RealLabs which hosts JetFuel, The Plug and other consumer apps, including Mango — the app published by Mango Labs on Google Play. Someone sure likes naming things “Labs!”)
Vungle claims the confusion here has to do with how it now uses the Mango Labs iTunes account to publish apps for its partners, which is a “common practice” on the App Store. It says it intends to transfer the apps published under Mango Labs to the developers’ accounts, because it agrees this is confusing.
Vungle also claims that JetFuel “does not make nor own any consumer apps that are currently live on the app stores. Any of the apps made by the entity when it was known as Mango Labs have long since been taken down from the app stores.”
JetFuel’s system is messy and confusing, but so far successful in its goals. Fontmaker did make it to No. 1, essentially growth hacked to the top by influencer marketing.
But as a consumer, what this all means is that you’ll never know who actually built the app you’re downloading or whether you were “influenced” to try it through what were, essentially, undisclosed ads.
Fontmaker isn’t the first to growth hack its way to the top through influencer promotions. Summertime hit Poparrazzi also hyped itself to the top of the App Store in a similar way, as have many others. But Poparazzi has since sunk to No. 89 in Photo & Video, which shows influence can only take you so far.
As for Fontmaker, paid influence got it to No. 1, but its Top Chart moment was brief.
Facebook is a monopoly. Right?
Mark Zuckerberg appeared on national TV today to make a “special announcement.” The timing could not be more curious: Today is the day Lina Khan’s FTC refiled its case to dismantle Facebook’s monopoly.
To the average person, Facebook’s monopoly seems obvious. “After all,” as James E. Boasberg of the U.S. District Court for the District of Columbia put it in his recent decision, “No one who hears the title of the 2010 film ‘The Social Network’ wonders which company it is about.” But obviousness is not an antitrust standard. Monopoly has a clear legal meaning, and thus far Lina Khan’s FTC has failed to meet it. Today’s refiling is much more substantive than the FTC’s first foray. But it’s still lacking some critical arguments. Here are some ideas from the front lines.
To the average person, Facebook’s monopoly seems obvious. But obviousness is not an antitrust standard.
First, the FTC must define the market correctly: personal social networking, which includes messaging. Second, the FTC must establish that Facebook controls over 60% of the market — the correct metric to establish this is revenue.
Though consumer harm is a well-known test of monopoly determination, our courts do not require the FTC to prove that Facebook harms consumers to win the case. As an alternative pleading, though, the government can present a compelling case that Facebook harms consumers by suppressing wages in the creator economy. If the creator economy is real, then the value of ads on Facebook’s services is generated through the fruits of creators’ labor; no one would watch the ads before videos or in between posts if the user-generated content was not there. Facebook has harmed consumers by suppressing creator wages.
A note: This is the first of a series on the Facebook monopoly. I am inspired by Cloudflare’s recent post explaining the impact of Amazon’s monopoly in their industry. Perhaps it was a competitive tactic, but I genuinely believe it more a patriotic duty: guideposts for legislators and regulators on a complex issue. My generation has watched with a combination of sadness and trepidation as legislators who barely use email question the leading technologists of our time about products that have long pervaded our lives in ways we don’t yet understand. I, personally, and my company both stand to gain little from this — but as a participant in the latest generation of social media upstarts, and as an American concerned for the future of our democracy, I feel a duty to try.
According to the court, the FTC must meet a two-part test: First, the FTC must define the market in which Facebook has monopoly power, established by the D.C. Circuit in Neumann v. Reinforced Earth Co. (1986). This is the market for personal social networking services, which includes messaging.
Second, the FTC must establish that Facebook controls a dominant share of that market, which courts have defined as 60% or above, established by the 3rd U.S. Circuit Court of Appeals in FTC v. AbbVie (2020). The right metric for this market share analysis is unequivocally revenue — daily active users (DAU) x average revenue per user (ARPU). And Facebook controls over 90%.
The answer to the FTC’s problem is hiding in plain sight: Snapchat’s investor presentations:
This is a chart of Facebook’s monopoly — 91% of the personal social networking market. The gray blob looks awfully like a vast oil deposit, successfully drilled by Facebook’s Standard Oil operations. Snapchat and Twitter are the small wildcatters, nearly irrelevant compared to Facebook’s scale. It should not be lost on any market observers that Facebook once tried to acquire both companies.
The market Includes messaging
The FTC initially claimed that Facebook has a monopoly of the “personal social networking services” market. The complaint excluded “mobile messaging” from Facebook’s market “because [messaging apps] (i) lack a ‘shared social space’ for interaction and (ii) do not employ a social graph to facilitate users’ finding and ‘friending’ other users they may know.”
This is incorrect because messaging is inextricable from Facebook’s power. Facebook demonstrated this with its WhatsApp acquisition, promotion of Messenger and prior attempts to buy Snapchat and Twitter. Any personal social networking service can expand its features — and Facebook’s moat is contingent on its control of messaging.
The more time in an ecosystem the more valuable it becomes. Value in social networks is calculated, depending on whom you ask, algorithmically (Metcalfe’s law) or logarithmically (Zipf’s law). Either way, in social networks, 1+1 is much more than 2.
Social networks become valuable based on the ever-increasing number of nodes, upon which companies can build more features. Zuckerberg coined the “social graph” to describe this relationship. The monopolies of Line, Kakao and WeChat in Japan, Korea and China prove this clearly. They began with messaging and expanded outward to become dominant personal social networking behemoths.
In today’s refiling, the FTC explains that Facebook, Instagram and Snapchat are all personal social networking services built on three key features:
- “First, personal social networking services are built on a social graph that maps the connections between users and their friends, family, and other personal connections.”
- “Second, personal social networking services include features that many users regularly employ to interact with personal connections and share their personal experiences in a shared social space, including in a one-to-many ‘broadcast’ format.”
- “Third, personal social networking services include features that allow users to find and connect with other users, to make it easier for each user to build and expand their set of personal connections.”
Unfortunately, this is only partially right. In social media’s treacherous waters, as the FTC has struggled to articulate, feature sets are routinely copied and cross-promoted. How can we forget Instagram’s copying of Snapchat’s stories? Facebook has ruthlessly copied features from the most successful apps on the market from inception. Its launch of a Clubhouse competitor called Live Audio Rooms is only the most recent example. Twitter and Snapchat are absolutely competitors to Facebook.
Messaging must be included to demonstrate Facebook’s breadth and voracious appetite to copy and destroy. WhatsApp and Messenger have over 2 billion and 1.3 billion users respectively. Given the ease of feature copying, a messaging service of WhatsApp’s scale could become a full-scale social network in a matter of months. This is precisely why Facebook acquired the company. Facebook’s breadth in social media services is remarkable. But the FTC needs to understand that messaging is a part of the market. And this acknowledgement would not hurt their case.
The metric: Revenue shows Facebook’s monopoly
Boasberg believes revenue is not an apt metric to calculate personal networking: “The overall revenues earned by PSN services cannot be the right metric for measuring market share here, as those revenues are all earned in a separate market — viz., the market for advertising.” He is confusing business model with market. Not all advertising is cut from the same cloth. In today’s refiling, the FTC correctly identifies “social advertising” as distinct from the “display advertising.”
But it goes off the deep end trying to avoid naming revenue as the distinguishing market share metric. Instead the FTC cites “time spent, daily active users (DAU), and monthly active users (MAU).” In a world where Facebook Blue and Instagram compete only with Snapchat, these metrics might bring Facebook Blue and Instagram combined over the 60% monopoly hurdle. But the FTC does not make a sufficiently convincing market definition argument to justify the choice of these metrics. Facebook should be compared to other personal social networking services such as Discord and Twitter — and their correct inclusion in the market would undermine the FTC’s choice of time spent or DAU/MAU.
Ultimately, cash is king. Revenue is what counts and what the FTC should emphasize. As Snapchat shows above, revenue in the personal social media industry is calculated by ARPU x DAU. The personal social media market is a different market from the entertainment social media market (where Facebook competes with YouTube, TikTok and Pinterest, among others). And this too is a separate market from the display search advertising market (Google). Not all advertising-based consumer technology is built the same. Again, advertising is a business model, not a market.
In the media world, for example, Netflix’s subscription revenue clearly competes in the same market as CBS’ advertising model. News Corp.’s acquisition of Facebook’s early competitor MySpace spoke volumes on the internet’s potential to disrupt and destroy traditional media advertising markets. Snapchat has chosen to pursue advertising, but incipient competitors like Discord are successfully growing using subscriptions. But their market share remains a pittance compared to Facebook.
An alternative pleading: Facebook’s market power suppresses wages in the creator economy
The FTC has correctly argued for the smallest possible market for their monopoly definition. Personal social networking, of which Facebook controls at least 80%, should not (in their strongest argument) include entertainment. This is the narrowest argument to make with the highest chance of success.
But they could choose to make a broader argument in the alternative, one that takes a bigger swing. As Lina Khan famously noted about Amazon in her 2017 note that began the New Brandeis movement, the traditional economic consumer harm test does not adequately address the harms posed by Big Tech. The harms are too abstract. As White House advisor Tim Wu argues in “The Curse of Bigness,” and Judge Boasberg acknowledges in his opinion, antitrust law does not hinge solely upon price effects. Facebook can be broken up without proving the negative impact of price effects.
However, Facebook has hurt consumers. Consumers are the workers whose labor constitutes Facebook’s value, and they’ve been underpaid. If you define personal networking to include entertainment, then YouTube is an instructive example. On both YouTube and Facebook properties, influencers can capture value by charging brands directly. That’s not what we’re talking about here; what matters is the percent of advertising revenue that is paid out to creators.
YouTube’s traditional percentage is 55%. YouTube announced it has paid $30 billion to creators and rights holders over the last three years. Let’s conservatively say that half of the money goes to rights holders; that means creators on average have earned $15 billion, which would mean $5 billion annually, a meaningful slice of YouTube’s $46 billion in revenue over that time. So in other words, YouTube paid creators a third of its revenue (this admittedly ignores YouTube’s non-advertising revenue).
Facebook, by comparison, announced just weeks ago a paltry $1 billion program over a year and change. Sure, creators may make some money from interstitial ads, but Facebook does not announce the percentage of revenue they hand to creators because it would be insulting. Over the equivalent three-year period of YouTube’s declaration, Facebook has generated $210 billion in revenue. one-third of this revenue paid to creators would represent $70 billion, or $23 billion a year.
Why hasn’t Facebook paid creators before? Because it hasn’t needed to do so. Facebook’s social graph is so large that creators must post there anyway — the scale afforded by success on Facebook Blue and Instagram allows creators to monetize through directly selling to brands. Facebooks ads have value because of creators’ labor; if the users did not generate content, the social graph would not exist. Creators deserve more than the scraps they generate on their own. Facebook suppresses creators’ wages because it can. This is what monopolies do.
Facebook’s Standard Oil ethos
Facebook has long been the Standard Oil of social media, using its core monopoly to begin its march upstream and down. Zuckerberg announced in July and renewed his focus today on the metaverse, a market Roblox has pioneered. After achieving a monopoly in personal social media and competing ably in entertainment social media and virtual reality, Facebook’s drilling continues. Yes, Facebook may be free, but its monopoly harms Americans by stifling creator wages. The antitrust laws dictate that consumer harm is not a necessary condition for proving a monopoly under the Sherman Act; monopolies in and of themselves are illegal. By refiling the correct market definition and marketshare, the FTC stands more than a chance. It should win.
A prior version of this article originally appeared on Substack.
After a judge slammed the F.T.C.’s original lawsuit, regulators returned with a more detailed case accusing the platform of being a monopoly.
Two Democratic senators have asked the new chair of the Federal Trade Commission to investigate Tesla’s statements about the autonomous capabilities of its Autopilot and Full Self-Driving systems. The senators, Edward Markey (D-Mass.) and Richard Blumenthal (D-Conn.), expressed particular concern over Tesla misleading customers into thinking their vehicles are capable of fully autonomous driving.
“Tesla’s marketing has repeatedly overstated the capabilities of its vehicles, and these statements increasingly pose a threat to motorists and other users of the road,” they said. “Accordingly, we urge you to open an investigation into potentially deceptive and unfair practices in Tesla’s advertising and marketing of its driving automation systems and take appropriate enforcement action to ensure the safety of all drivers on the road.”
In their letter to new FTC Chair Lina Khan, they point to a 2019 Youtube video Tesla posted to its channel, which shows a Tesla driving autonomously. The roughly two-minute-long video is titled “Full Self-Driving” and has been viewed over 18 million times.
“Their claims put Tesla drivers – and all of the travelling public – at risk of serious injury or death,” the Senators said.
When it comes to Tesla and formal investigations, when it rains, it pours. The letter was published just two days after the National Highway Transportation Safety Administration said it had opened a preliminary investigation into incidents involving Teslas crashing into parked emergency vehicles.
Lina Khan is the youngest person to ever chair the FTC. She’s widely considered the most progressive appointment in recent history, particularly for her scholarship on antitrust law. But should the FTC choose to investigate Tesla, the case would likely have nothing to do with antitrust law and instead fall under the purview of consumer protection. The FTC has the authority to investigate false or misleading claims from companies regarding their products.
This is not the first time prominent figures have called on the FTC to open an investigation into Tesla’s claims. The Center for Auto Safety and Consumer Watchdog, two special interest groups, also sent a letter in 2018 to the commission over the marketing of Autopilot features. The following year, the NHTSA urged the FTC to investigate whether claims made by Tesla CEO Elon Musk on the Model 3’s safety “constitute[d] unfair or deceptive acts or practices.”
Tesla charges $10,000 for access to a “Full Self-Driving” option at the point of sale, or as a subscription. The company is currently testing beta version 9 of FSD with a few thousand drivers, but the Senators take aim at the beta version, too. “After the [beta 9] update, drivers have posted videos online showing their updated Tesla vehicles making unexpected maneuvers that require human intervention to prevent a crash,” they write. “Mr. Musk’s tepid precautions tucked away on social media are no excuse for misleading drivers and endangering the lives of everyone on the road.”
Advocates of algorithmic justice have begun to see their proverbial “days in court” with legal investigations of enterprises like UHG and Apple Card. The Apple Card case is a strong example of how current anti-discrimination laws fall short of the fast pace of scientific research in the emerging field of quantifiable fairness.
While it may be true that Apple and their underwriters were found innocent of fair lending violations, the ruling came with clear caveats that should be a warning sign to enterprises using machine learning within any regulated space. Unless executives begin to take algorithmic fairness more seriously, their days ahead will be full of legal challenges and reputational damage.
What happened with Apple Card?
In late 2019, startup leader and social media celebrity David Heinemeier Hansson raised an important issue on Twitter, to much fanfare and applause. With almost 50,000 likes and retweets, he asked Apple and their underwriting partner, Goldman Sachs, to explain why he and his wife, who share the same financial ability, would be granted different credit limits. To many in the field of algorithmic fairness, it was a watershed moment to see the issues we advocate go mainstream, culminating in an inquiry from the NY Department of Financial Services (DFS).
At first glance, it may seem heartening to credit underwriters that the DFS concluded in March that Goldman’s underwriting algorithm did not violate the strict rules of financial access created in 1974 to protect women and minorities from lending discrimination. While disappointing to activists, this result was not surprising to those of us working closely with data teams in finance.
There are some algorithmic applications for financial institutions where the risks of experimentation far outweigh any benefit, and credit underwriting is one of them. We could have predicted that Goldman would be found innocent, because the laws for fairness in lending (if outdated) are clear and strictly enforced.
And yet, there is no doubt in my mind that the Goldman/Apple algorithm discriminates, along with every other credit scoring and underwriting algorithm on the market today. Nor do I doubt that these algorithms would fall apart if researchers were ever granted access to the models and data we would need to validate this claim. I know this because the NY DFS partially released its methodology for vetting the Goldman algorithm, and as you might expect, their audit fell far short of the standards held by modern algorithm auditors today.
How did DFS (under current law) assess the fairness of Apple Card?
In order to prove the Apple algorithm was “fair,” DFS considered first whether Goldman had used “prohibited characteristics” of potential applicants like gender or marital status. This one was easy for Goldman to pass — they don’t include race, gender or marital status as an input to the model. However, we’ve known for years now that some model features can act as “proxies” for protected classes.
If you’re Black, a woman and pregnant, for instance, your likelihood of obtaining credit may be lower than the average of the outcomes among each overarching protected category.
The DFS methodology, based on 50 years of legal precedent, failed to mention whether they considered this question, but we can guess that they did not. Because if they had, they’d have quickly found that credit score is so tightly correlated to race that some states are considering banning its use for casualty insurance. Proxy features have only stepped into the research spotlight recently, giving us our first example of how science has outpaced regulation.
In the absence of protected features, DFS then looked for credit profiles that were similar in content but belonged to people of different protected classes. In a certain imprecise sense, they sought to find out what would happen to the credit decision were we to “flip” the gender on the application. Would a female version of the male applicant receive the same treatment?
Intuitively, this seems like one way to define “fair.” And it is — in the field of machine learning fairness, there is a concept called a “flip test” and it is one of many measures of a concept called “individual fairness,” which is exactly what it sounds like. I asked Patrick Hall, principal scientist at bnh.ai, a leading boutique AI law firm, about the analysis most common in investigating fair lending cases. Referring to the methods DFS used to audit Apple Card, he called it basic regression, or “a 1970s version of the flip test,” bringing us example number two of our insufficient laws.
A new vocabulary for algorithmic fairness
Ever since Solon Barocas’ seminal paper “Big Data’s Disparate Impact” in 2016, researchers have been hard at work to define core philosophical concepts into mathematical terms. Several conferences have sprung into existence, with new fairness tracks emerging at the most notable AI events. The field is in a period of hypergrowth, where the law has as of yet failed to keep pace. But just like what happened to the cybersecurity industry, this legal reprieve won’t last forever.
Perhaps we can forgive DFS for its softball audit given that the laws governing fair lending are born of the civil rights movement and have not evolved much in the 50-plus years since inception. The legal precedents were set long before machine learning fairness research really took off. If DFS had been appropriately equipped to deal with the challenge of evaluating the fairness of the Apple Card, they would have used the robust vocabulary for algorithmic assessment that’s blossomed over the last five years.
The DFS report, for instance, makes no mention of measuring “equalized odds,” a notorious line of inquiry first made famous in 2018 by Joy Buolamwini, Timnit Gebru and Deb Raji. Their “Gender Shades” paper proved that facial recognition algorithms guess wrong on dark female faces more often than they do on subjects with lighter skin, and this reasoning holds true for many applications of prediction beyond computer vision alone.
Equalized odds would ask of Apple’s algorithm: Just how often does it predict creditworthiness correctly? How often does it guess wrong? Are there disparities in these error rates among people of different genders, races or disability status? According to Hall, these measurements are important, but simply too new to have been fully codified into the legal system.
If it turns out that Goldman regularly underestimates female applicants in the real world, or assigns interest rates that are higher than Black applicants truly deserve, it’s easy to see how this would harm these underserved populations at national scale.
Financial services’ Catch-22
Modern auditors know that the methods dictated by legal precedent fail to catch nuances in fairness for intersectional combinations within minority categories — a problem that’s exacerbated by the complexity of machine learning models. If you’re Black, a woman and pregnant, for instance, your likelihood of obtaining credit may be lower than the average of the outcomes among each overarching protected category.
These underrepresented groups may never benefit from a holistic audit of the system without special attention paid to their uniqueness, given that the sample size of minorities is by definition a smaller number in the set. This is why modern auditors prefer “fairness through awareness” approaches that allow us to measure results with explicit knowledge of the demographics of the individuals in each group.
But there’s a Catch-22. In financial services and other highly regulated fields, auditors often can’t use “fairness through awareness,” because they may be prevented from collecting sensitive information from the start. The goal of this legal constraint was to prevent lenders from discrimination. In a cruel twist of fate, this gives cover to algorithmic discrimination, giving us our third example of legal insufficiency.
The fact that we can’t collect this information hamstrings our ability to find out how models treat underserved groups. Without it, we might never prove what we know to be true in practice — full-time moms, for instance, will reliably have thinner credit files, because they don’t execute every credit-based purchase under both spousal names. Minority groups may be far more likely to be gig workers, tipped employees or participate in cash-based industries, leading to commonalities among their income profiles that prove less common for the majority.
Importantly, these differences on the applicants’ credit files do not necessarily translate to true financial responsibility or creditworthiness. If it’s your goal to predict creditworthiness accurately, you’d want to know where the method (e.g., a credit score) breaks down.
What this means for businesses using AI
In Apple’s example, it’s worth mentioning a hopeful epilogue to the story where Apple made a consequential update to their credit policy to combat the discrimination that is protected by our antiquated laws. In Apple CEO Tim Cook’s announcement, he was quick to highlight a “lack of fairness in the way the industry [calculates] credit scores.”
Their new policy allows spouses or parents to combine credit files such that the weaker credit file can benefit from the stronger. It’s a great example of a company thinking ahead to steps that may actually reduce the discrimination that exists structurally in our world. In updating their policies, Apple got ahead of the regulation that may come as a result of this inquiry.
This is a strategic advantage for Apple, because NY DFS made exhaustive mention of the insufficiency of current laws governing this space, meaning updates to regulation may be nearer than many think. To quote Superintendent of Financial Services Linda A. Lacewell: “The use of credit scoring in its current form and laws and regulations barring discrimination in lending are in need of strengthening and modernization.” In my own experience working with regulators, this is something today’s authorities are very keen to explore.
I have no doubt that American regulators are working to improve the laws that govern AI, taking advantage of this robust vocabulary for equality in automation and math. The Federal Reserve, OCC, CFPB, FTC and Congress are all eager to address algorithmic discrimination, even if their pace is slow.
In the meantime, we have every reason to believe that algorithmic discrimination is rampant, largely because the industry has also been slow to adopt the language of academia that the last few years have brought. Little excuse remains for enterprises failing to take advantage of this new field of fairness, and to root out the predictive discrimination that is in some ways guaranteed. And the EU agrees, with draft laws that apply specifically to AI that are set to be adopted some time in the next two years.
The field of machine learning fairness has matured quickly, with new techniques discovered every year and myriad tools to help. The field is only now reaching a point where this can be prescribed with some degree of automation. Standards bodies have stepped in to provide guidance to lower the frequency and severity of these issues, even if American law is slow to adopt.
Because whether discrimination by algorithm is intentional, it is illegal. So, anyone using advanced analytics for applications relating to healthcare, housing, hiring, financial services, education or government are likely breaking these laws without knowing it.
Until clearer regulatory guidance becomes available for the myriad applications of AI in sensitive situations, the industry is on its own to figure out which definitions of fairness are best.
Facebook’s decision to close accounts connected to a misinformation research project last week prompted a broad outcry from the company’s critics — and now Congress is getting involved.
A handful of lawmakers criticized the decision at the time, slamming Facebook for being hostile toward efforts to make the platform’s opaque algorithms and ad targeting methods more transparent. Researchers believe that studying those hidden systems is crucial work for gaining insight on the flow of political misinformation.
The company specifically punished two researchers with NYU’s Cybersecurity for Democracy project who work on Ad Observer, an opt-in browser tool that allows researchers to study how Facebook targets ads to different people based on their interests and demographics.
In a new letter, embedded below, a trio of Democratic senators are pressing Facebook for more answers. Senators Amy Klobuchar (D-MN), Chris Coons (D-DE) and Mark Warner (D-VA) wrote to Facebook CEO Mark Zuckerberg asking for a full explanation on why the company terminated the researcher accounts and how they violated the platform’s terms of service and compromised user privacy. The lawmakers sent the letter on Friday.
“While we agree that Facebook must safeguard user privacy, it is similarly imperative that Facebook allow credible academic researchers and journalists like those involved in the Ad Observatory project to conduct independent research that will help illuminate how the company can better tackle misinformation, disinformation, and other harmful activity that is proliferating on its platforms,” the senators wrote.
Lawmakers have long urged the company to be more transparent about political advertising and misinformation, particularly after Facebook was found to have distributed election disinformation in 2016. Those concerns were only heightened by the platform’s substantial role in spreading election misinformation leading up to the insurrection at the U.S. Capitol, where Trump supporters attempted to overturn the vote.
In a blog post defending its decision, Facebook cited compliance with FTC as one of the reason the company severed the accounts. But the FTC called Facebook’s bluff last week in a letter to Zuckerberg, noting that nothing about the agency’s guidance for the company would preclude it from encouraging research in the public interest.
“Indeed, the FTC supports efforts to shed light on opaque business practices, especially around surveillance-based advertising,” Samuel Levine, the FTC’s acting director for the Bureau of Consumer Protection, wrote.
Facebook shut down accounts belonging to two academic researchers late Tuesday, cutting off their ability to study political ads and misinformation on the world’s biggest social network.
The company accused the academics of engaging in “unauthorized scraping” and compromising user privacy on the platform, claims that Facebook’s many critics are slamming as a thin pretense for killing the transparency work.
The company took action against Laura Edelson and Damon McCoy, two well-known researchers affiliated with NYU’s Cybersecurity for Democracy project who have long sparred with the company. The move cuts off their access to Facebook’s Ad Library — one of the company’s only meaningful transparency efforts to date — and data on popular posts from the social media monitoring service CrowdTangle.
Facebook has a history with Edelson and McCoy. The company served the pair cease and desist letters just weeks before the 2020 election, calling on the team to disable an opt-in browser tool called Ad Observer and unpublish their findings. Ad Observer is a browser tool anyone can install that’s designed to give researchers a rare glimpse into how Facebook targets the ads that have transformed it into a trillion-dollar company.
“Over the last several years, we’ve used this access to uncover systemic flaws in the Facebook Ad Library, identify misinformation in political ads including many sowing distrust in our election system, and to study Facebook’s apparent amplification of partisan misinformation,” Edelson said on Twitter.
“By suspending our accounts, Facebook has effectively ended all this work. Facebook has also effectively cut off access to more than two dozen other researchers and journalists who get access to Facebook data through our project, including our work measuring vaccine misinformation with the Virality Project and many other partners who rely on our data.”
The incident set off a fresh round of criticism about the company’s preference for opacity over transparency when it comes to some of the more dangerous behavior that the platform incubates.
By Wednesday, Facebook’s actions had attracted the attention of some members of Congress. Sen. Ron Wyden (D-OR) criticized Facebook’s decision to punish the researchers under the pretense of protecting users in light of the company’s long history of invasive privacy practices. Wyden also called Facebook’s bluff over its claim that revoking researcher access is an effort to comply with a privacy order from the FTC that the company was issued for its previous user privacy violations.
Sen. Mark Warner (D-VA) also weighed in on Facebook’s latest controversy, calling the decision “deeply concerning.” Warner praised independent researchers for “consistently [improving] the integrity and safety of social media platforms by exposing harmful and exploitative activity.”
“It’s past time for Congress to act to bring greater transparency to the shadowy world of online advertising, which continues to be a major vector for fraud and misconduct,” Warner said.
A number of free press organizations, researchers and misinformation experts also condemned Facebook’s decision Wednesday. “Facebook’s cavalier approach to privacy enabled it to become so dominant,” The Markup’s Julia Angwin and Nabiha Syed wrote in a joint statement.
“But now, when independent researchers want to interrogate that platform and the influence it commands, Facebook is propping up user privacy as a shield to hide behind.”
The Biden administration tripled down on its commitment to reining in powerful tech companies Tuesday, proposing committed Big Tech critic Jonathan Kanter to lead the Justice Department’s antitrust division.
Kanter is a lawyer with a long track record of representing smaller companies like Yelp in antitrust cases against Google. He currently practices law at his own firm, which specializes in advocacy for state and federal antitrust enforcement.
“Throughout his career, Kanter has also been a leading advocate and expert in the effort to promote strong and meaningful antitrust enforcement and competition policy,” the White House press release stated. Progressives celebrated the nomination as a win, though some of Biden’s new antitrust hawks have enjoyed support from both political parties.
The Justice Department already has a major antitrust suit against Google in the works. The lawsuit, filed by Trump’s own Justice Department, accuses the company of “unlawfully maintaining monopolies” through anti-competitive practices in its search and search advertising businesses. If successfully confirmed, Kanter would be positioned to steer the DOJ’s big case against Google.
In a 2016 NYT op-ed, Kanter argued that Google is notorious for relying on an anti-competitive “playbook” to maintain its market dominance. Kanter pointed to Google’s long history of releasing free ad-supported products and eventually restricting competition through “discriminatory and exclusionary practices” in a given corner of the market.
Kanter is just the latest high profile Big Tech critic that’s been elevated to a major regulatory role under Biden. Last month, Biden named fierce Amazon critic Lina Khan as FTC chair upon her confirmation to the agency. In March, Biden named another noted Big Tech critic, Columbia law professor Tim Wu, to the National Economic Council as a special assistant for tech and competition policy.
All signs point to the Biden White House gearing up for a major federal fight with Big Tech. Congress is working on a set of Big Tech bills, but in lieu of — or in tandem with — legislative reform, the White House can flex its own regulatory muscle through the FTC and DOJ.
In new comments to MSNBC, the White House confirmed that it is also “reviewing” Section 230 of the Communications Decency Act, a potent snippet of law that protects platforms from liability for user-generated content.
PayPal-owned payments app Venmo will no longer offer a public, global feed of users’ transactions, as part of a significant redesign focused on expanding the app’s privacy controls and better highlighting some of Venmo’s newer features. The company says it will instead only show users their “friends feed” — meaning, the app’s social feed where you can see just your friends’ transactions.
Venmo has struggled over the years to balance its desire to add a social element to its peer-to-peer payments-based network, with the need to offer users their privacy.
A few years ago, the company was forced to settle a complaint with the FTC over its handling of privacy disclosures in the app along with other issues related to the security and privacy of user transactions. One of the concerns at the time was a setting that made all transactions public by default — a feature the FTC said wasn’t being properly explained to customers. As part of the settlement, Venmo had to inform both new and existing users how to limit the visibility of their transactions, among other changes.
However, privacy issues have continued to follow Venmo over the years. More recently, BuzzFeed News was able to track down President Biden’s secret Venmo account because of the lack of privacy around Venmo friend lists, for example. Afterwards, the company rolled out friend list privacy controls to address the issue.
In the newly updated app, Venmo will still highlight this friend list privacy setting so users can choose whether or not they want to have their profile appear on other people’s friends’ lists. Users will also still be able to remove or add contacts from their friend list at any time, block people, and set their transaction privacy either as they post or retroactively to public, private, or friends-only. It’s unclear what advantage posting publicly has though, as the global, public feed is gone. Instead, public transactions would be visible to a users’ non-friends only when someone visited their profile directly.
In addition to the privacy changes, Venmo’s redesign aims to make it easier for people to discover the app’s new features, the company says.
Now, a new bottom navigation option will allow users to toggle between their social feed, Venmo’s products like the Venmo Card and crypto, and their personal profile. The newly elevated “Cards” section will allow Venmo Credit and Debit cardholders to manage their cards and access their rewards and offers, as before. Meanwhile, the “Crypto” tab will let users learn and explore the world of crypto, view real-time trends, and buy, sell or hold different types of cryptocurrencies.
Venmo first added support for crypto earlier this year, following parent company PayPal’s move to do the same, and now offers access to Bitcoin, Ethereum, Litecoin and Bitcoin Cash. Before, the option appeared as a small button next to the “Pay or Request” button at the bottom of the screen, which contributed to Venmo’s cluttered feel.
The updated app will also include support for new payment types and expanded purchase protections, which Venmo announced last month, and said would arrive on July 20. Customers will now be able to indicate if their purchase is for “goods and services” when they transact with a seller, which will make the transactions eligible for Venmo’s purchase protection plan — even if the seller doesn’t have a proper “business” account.
Because this now charges sellers a 1.9% plus 10-cent fee, there had been some backlash from users who either misunderstood the changes or just didn’t like them. But the move could help to boost Venmo revenue.
PayPal said in February that Venmo grew users 32% over 2020 to reach 70 million active accounts and expects the app to generate nearly $900 million in revenue this year — likely in part thanks to this and other new initiatives, like its crypto transaction fees.
Beyond the more functional changes and the privacy updates, Venmo’s redesign also modernizes the look-and-feel of the app itself, which had become a little dated and overly busy. As Venmo had expanded its array of services, the hamburger (three line) menu in the top right of the old version of the app had turned into a long list of options and settings. Now that’s gone. The app uses new iconography, an updated font, and lots of white space to make it feel fresh and clean.
The app’s changes also somewhat de-emphasize the importance of the social feed itself. Although it may still default to that tab, other options now have equal footing with tabs of their own, instead of being hidden away in a menu or in a smaller button.
Venmo says the redesigned Venmo app will begin to roll out today to select customers and will be available to all users across the U.S. over the next few weeks.
An F.C.C. rule that went into effect last month is meant to help put a stop to those relentless calls about your extended warranty, and others.
The Biden administration just introduced a sweeping, ambitious plan to forcibly inject competition into some consolidated sectors of the American economy — the tech sector prominent among them — through executive action.
“Today President Biden is taking decisive action to reduce the trend of corporate consolidation, increase competition, and deliver concrete benefits to America’s consumers, workers, farmers, and small businesses,” a new White House fact sheet on the forthcoming order states.
The order, which Biden will sign Friday, initiates a comprehensive “whole-of-government” approach that loops in more then twelve different agencies at the federal level to regulate monopolies, protect consumers and curtail bad behavior from some of the world’s biggest corporations.
In the fact sheet, the White House lays out its plans to take matters to regulate big business into its own hands at the federal level. As far as tech is concerned, that comes largely through emboldening the FTC and the Justice Department — two federal agencies with antitrust enforcement powers.
Most notably for big tech, which is already bracing for regulatory existential threats, the White House explicitly asserts here that those agencies have legal cover to “challenge prior bad mergers that past Administrations did not previously challenge” — i.e. unwinding acquisitions that built a handful of tech companies into the behemoths they are today. The order calls on antitrust agencies to enforce antitrust laws “vigorously.”
Federal scrutiny will prioritize “dominant internet platforms, with particular attention to the acquisition of nascent competitors, serial mergers, the accumulation of data, competition by ‘free’ products, and the effect on user privacy.” Facebook, Google and Amazon are particularly on notice here, though Apple isn’t likely to escape federal attention either.
“Over the past ten years, the largest tech platforms have acquired hundreds of companies—including alleged ‘killer acquisitions’ meant to shut down a potential competitive threat,” the White House wrote in the fact sheet. “Too often, federal agencies have not blocked, conditioned, or, in some cases, meaningfully examined these acquisitions.”
The biggest tech companies have regularly defended their longstanding strategy of buying up the competition by arguing that because those acquisitions went through without friction at the time, they shouldn’t be viewed as illegal in hindsight. In no uncertain terms, the new executive order makes it clear that the Biden administration isn’t having any of it.
The White House also specifically singles out internet service providers for scrutiny, ordering the FCC to prioritize consumer choice and institute broadband “nutrition labels” that clearly state speed caps and hidden feeds. The FCC began working on the labels in the Obama administration but the work was scrapped after Trump took office.
The order also directly calls on the FCC to restore net neutrality rules, which were stripped in 2017 to the widespread horror of open internet advocates and most of the tech industry outside of the service providers that stood to benefit.
The White House will also tell the FTC to create new privacy rules meant to guard consumers against surveillance and the “accumulation of extraordinarily amounts of sensitive personal information,” which free services like Facebook, YouTube and others have leveraged to build their vast empires. The White House also taps the FTC to create rules that protect smaller businesses from being pre-empted by large platforms, which in many cases abuse their market dominance with a different sort of data-based surveillance to out-compete up-and-coming competitors.
Finally, the executive order encourages the FTC to put right to repair rules in place that would free consumers from constraints that discourage DIY and third-party repairs. A new White House Competition Council under the Director of the National Economic Council will coordinate the federal execution of the proposals laid out in the new order.
The antitrust effort from the executive branch mirrors parallel actions in the FTC and Congress. In the FTC, Biden has installed a fearsome antitrust crusader in Lina Khan, a young legal scholar and fierce Amazon critic who proposes a philosophical overhaul to the way the federal government defines monopolies. Khan now leads the FTC as its chair.
In Congress, a bipartisan flurry of bills intended to rein in the tech industry are slowly wending their way toward becoming law, though plenty of hurdles remain. Last month, the House Judiciary Committee debated the six bills, which were crafted separately to help them survive opposing lobbying pushes from the tech industry. These legislative efforts could modernize antitrust laws, which have failed to keep pace with the modern realities of giant, internet-based businesses.
“Competition policy needs new energy and approaches so that we can address America’s monopoly problem,” Sen. Amy Klobuchar, a prominent tech antitrust hawk in Congress, said of the executive order. “That means legislation to update our antitrust laws, but it also means reimagining what the federal government can do to promote competition under our current laws.”
Citing the acceleration of corporate consolidation in recent decades, the White House argues that a handful of large corporations dominates across industries, including healthcare, agriculture and tech and consumers, workers and smaller competitors pay the price for their outsized success. The administration will focus antitrust enforcement on those corners of the market as well as evaluating the labor market and worker protections on the whole.
“Inadequate competition holds back economic growth and innovation… Economists find that as competition declines, productivity growth slows, business investment and innovation decline, and income, wealth, and racial inequality widen,” the White House wrote.
An executive order reflects the administration’s growing embrace of warnings by some economists that declining competition is hobbling the economy’s vitality.
The order is the president’s latest acknowledgment of concerns that the tech giants have obtained outsize market power.
The company faces declining sales and thousands of lawsuits claiming it knowingly sold its trendy vaping products to minors. Soon the F.D.A. will decide whether it can keep selling them at all.