program areas Digital Citizen

The Trade Desk is a leading (link is external) AdTech company, providing data-driven digital advertising services (link is external) to major brands and agencies. It is also playing an outsized role responding to the initiative led by Google (link is external) to create new, allegedly “privacy-friendly” approaches to ad targeting, which include ending the use of what are called “third-party” cookies. These cookies enable the identification and tracking of individuals, and have been an essential building block for surveillance advertising since the dawn (link is external) of the commercial Internet. As we explained in a previous post about the so-called race to “end” the use of cookies, the online marketing industry is engaged in a full-throated effort to redefine how our privacy is conceptualized and privately governed. Pressure from regulators (such as the EU’s GDPR) and growing concerns about privacy from consumers are among the reasons why this is happening now. But the real motivation, in my view, is that the most powerful online ad companies and global brands (such as Google, Amazon and the Trade Desk) don’t need these antiquated cookies anymore. They have so much of our information that they collect directly, and also available from countless partners (such as global brands). Additionally, they now have many new ways to determine who we are—our “identity”—including through the use of AI, machine learning and data clouds (link is external). “Unified ID 2.0” is what The Trade Desk calls its approach to harvesting our identity information for advertising. Like Google, they claim to be respectful of data protection principles. Some of the most powerful companies in the U.S. are supporting the Unified ID standard, including Walmart, Washington Post, P&G, Comcast, CBS, Home Depot, Oracle, and Nielsen. But more than our privacy is at stake as data marketing giants fight over how best to reap the financial rewards (link is external) of what is predicted eventually to become a trillion dollar global ad marketplace. This debate is increasingly focused on the very future of the Internet itself, including how it is structured and governed. Only by ensuring that advertisers can continue to successfully operate powerful data-gathering and ad-targeting systems, argues Trade Desk CEO Jeff Green, can the “Open (link is external) Internet” be preserved. His argument, of course, is a digital déjà vu version of what media moguls have said in the U.S. dating back to commercial radio in the 1930’s. Only with a full-blown, ad-supported (and regulation-free) electronic media system, whether it was broadcast radio, broadcast TV, or cable TV, could the U.S. be assured it would enjoy a democratic and robust communications environment. (I was in the room at the Department of Commerce back in the middle 1990’s when advertisers were actually worried that the Internet would be largely ad-free; the representative from P&G leaned over to tell me that they never would let that happen—and he was right.) Internet operations are highly influenced to serve the needs of advertisers, who have reworked its architecture to ensure we are all commercially surveilled. For decades, the online ad industry has continually expanded ways to monetize our behaviors, emotions, location and much more. (link is external) Last week, The Trade Desk unveiled its latest iteration using Unified ID 2.0—called Solimar (see video (link is external) here). Solimar uses “an artificial intelligence tool called Koa (link is external), which makes suggestions” to help ensure effective marketing campaigns. Reflecting the serial partnerships that operate to provide marketers with a gold mine of information on any individual, The Trade Desk has a “Koa Identity (link is external) Alliance,” a “cross-device graph that incorporates leading and emerging ID solutions such as LiveRamp Identity Link, Oracle Cross Device, Tapad (link is external) Device Graph, and Adbrain Device Graf.” This system, they say, creates an effective way for marketers to develop a data portrait of individual consumers. It’s useful to hear what companies such as The Trade Desk say as we evaluate claims that “big data” consumer surveillance operations are essential for a democratically structured Internet. In its most recent Annual Report (link is external), the company explains that “Through our self-service, cloud-based platform, ad buyers can create, manage, and optimize more expressive data-driven digital advertising campaigns across ad formats and channels, including display, video, audio, in-app, native and social, on a multitude of devices, such as computers, mobile devices, and connected TV (‘CTV’)…. We use the massive data captured by our platform to build predictive models around user characteristics, such as demographic, purchase intent or interest data. Data from our platform is continually fed back into these models, which enables them to improve over time as the use of our platform increases.” And here’s how The Trade Desk’s Koa’s process is described in the trade publication Campaign (link is external) Asia: …clients can specify their target customer in the form of first-party or third-party data, which will serve as a seed audience that Koa will model from to provide recommendations. A data section provides multiple options for brands to upload first-party data including pixels, app data, and IP addresses directly into the platform, or import data from a third-party DMP or CDP. If a client chooses to onboard CRM data in the form of email addresses, these will automatically be converted into UID2s. Once converted, the platform will scan the UID2s to evaluate how many are ‘active UID2s’, which refers to how many of these users have been active across the programmatic universe in the past week. If the client chooses to act on those UID2s, they will be passed into the programmatic ecosystem to match with the publisher side, building the UID2 ecosystem in tandem. For advertisers that don't have first-party data… an audiences tab allows advertisers to tap into a marketplace of second- and third-party data so they can still use interest segments, purchase intent segments and demographics. In other words, these systems have a ton of information about you. They can easily get even more data and engage in the kinds of surveillance advertising that regulators (link is external) and consumer (link is external) advocates around the world are demanding be stopped. There are now dozens of competing “identity solutions”—including those from Google, Amazon (link is external), data brokers (link is external), telephone (link is external) companies, etc. (See visual at bottom of page here (link is external)). The stakes here are significant—how will the Internet evolve in terms of privacy, and will its core “DNA” be ever-growing forms of surveillance and manipulation? How do we decide the most privacy-protective ways to ensure meaningful monetization of online content—and must funding for such programming only be advertising-based? In what ways are some of these identity proposals a way for powerful platforms such as Google to further expand its monopolistic control of the ad market? These and other questions require a thoughtful regulator in the U.S. to help sort this out and make recommendations to ensure that the public truly benefits. That’s why it’s time for the U.S. Federal Trade Commission to step in. The FTC should analyze these advertising-focused identity efforts; assess their risks and the benefits; address how to govern the collection and use of data where a person has supposedly given permission to a brand or store to use it (known as “first-party” data). A key question, given today’s technologies, is whether meaningful personal consent for data collection is even possible in a world driven by sophisticated and real-time AI systems that personalize content and ads? The commission should also investigate the role of data-mining clouds and other so-called “clean” rooms where privacy is said to prevail despite their compilation of personal information for targeted advertising. The time for private, special interests (and conflicted) actors to determine the future of the Internet, and how our privacy is to be treated, is over.

International coalition calls for action against surveillance-based advertising Every day, consumers are exposed to extensive commercial surveillance online. This leads to manipulation, fraud, discrimination and privacy violations. Information about what we like, our purchases, mental and physical health, sexual orientation, location and political views are collected, combined and used under the guise of targeting advertising.   In a new report, the Norwegian Consumer Council (NCC) sheds light on the negative consequences that this commercial surveillance has on consumers and society. Together with [XXX] organizations and experts, NCC is asking authorities on both sides of the Atlantic to consider a ban. In Europe, the upcoming Digital Services Act can lay the legal framework to do so. In the US, legislators should seize the opportunity to enact comprehensive privacy legislation that protects consumers.  - The collection and combination of information about us not only violates our right to privacy, but renders us vulnerable to manipulation, discrimination and fraud. This harms individuals and society as a whole, says the director of digital policy in the NCC, Finn Myrstad.  In a Norwegian population survey conducted by YouGov on behalf of the NCC, consumers clearly state that they do not want commercial surveillance. Just one out of ten respondents were positive to commercial actors collecting personal information about them online, while only one out of five thought that ads based on personal information is acceptable. - Most of us do not want to be spied on online, or receive ads based on tracking and profiling. These results mirror similar surveys from Europe and the United States, and should be a powerful signal to policymakers looking at how to better regulate the internet, Myrstad says. Policymakers and civil society organisations on both sides of the Atlantic are increasingly standing up against these invasive practices. For example, The European Parliament and the European Data Protection Supervisor (EDPS) have already called for phasing out and banning surveillance-based advertising. A coalition of consumer and civil rights organizations in the United States has called for a similar ban.     Significant consequences  The NCC report ’Time to ban surveillance-based advertising’ exposes a variety of harmful consequences that surveillance-based advertising can have on individuals and on society:    1. Manipulation  Companies with comprehensive and intimate knowledge about us can shape their messages in attempts to reach us when we are susceptible, for example to influence elections or to advertise weight loss products, unhealthy food or gambling.     2. Discrimination  The opacity and automation of surveillance-based advertising systems increase the risk of discrimination, for example by excluding consumers based on income, gender, race, ethnicity or sexual orientation, location, or by making certain consumers pay more for products or services.     3. Misinformation   The lack of control over where ads are shown can promote and finance false or malicious content. This also poses significant challenges to publishers and advertisers regarding revenue, reputational damage, and opaque supply chains. 4. Undermining competition   The surveillance business model favours companies that collect and process information across different services and platforms. This makes it difficult for smaller actors to compete, and negatively impacts companies that respect consumers’ fundamental rights.  5. Security risks  When thousands of companies collect and process enormous amounts of personal data, the risk of identity theft, fraud and blackmail increases. NATO has described this data collection as a national security risk.    6. Privacy violations   The collection and use of personal data is happening with little or no control, both by large companies and by companies that are unknown to most consumers. Consumers have no way to know what data is collected, who the information is shared with, and how it may be used.   -  It is very difficult to justfy the negative consequences of this system. A ban will contribute to a healthier marketplace that helps protect individuals and society, Myrstad comments.  Good alternatives  In the report, the NCC points to alternative digital advertising models that do not depend on the surveillance of consumers, and that provide advertisers and publishers more oversight and control over where ads are displayed and which ads are being shown. - It is possible to sell advertising space without basing it on intimate details about consumers. Solutions already exist to show ads in relevant contexts, or where consumers self-report what ads they want to see, Myrstad says. - A ban on surveillance-based advertising would also pave the way for a more transparent advertising marketplace, diminishing the need to share large parts of ad revenue with third parties such as data brokers. A level playing field would contribute to giving advertisers and content providers more control, and keep a larger share of the revenue. The coordinated push behind the report and letter illustrates the growing determination of consumer, digital rights, human rights and other civil society groups to end the widespread business model of spying on the public.

The Whole World will Still be Watching You: Google & Digital Marketing Industry “Death-of-the-Cookie” Privacy Initiatives Require Scrutiny from Public Policymakers Jeff Chester One would think, in listening to the language used by Google, Facebook, and other ad and data companies to discuss the construction and future of privacy protection, that they are playing some kind of word game. We hear terms (link is external) such as “TURTLEDOVE,” “FLEDGE,” SPARROW and “FLoC.” Such claims should be viewed with skepticism, however. Although some reports make it appear that Google and its online marketing compatriots propose to reduce data gathering and tracking, we believe that their primary goal is still focused on perfecting the vast surveillance system they’ve well-established. A major data marketing industry effort is now underway to eliminate—or diminish—the role of the tracking software known as “third-party” cookies. Cookies were developed (link is external) in the very earliest days of the commercial “World Wide Web,” and have served as the foundational digital tether connecting us to a sprawling and sophisticated data-mining complex. Through cookies—and later mobile device IDs and other “persistent” identifiers—Google, Facebook, Amazon, Coca-Cola and practically everyone else have been able to surveil and target us—and our communities. Tracking cookies have literally helped engineer a “sweet spot (link is external)” for online marketers, enabling them to embed spies into our web browsers, which help them understand our digital behaviors and activities and then take action based on that knowledge. Some of these trackers—placed and used by a myriad (link is external) of data marketing companies on various websites—are referred to as “third-party” cookies, to distinguish them from what online marketers claim, with a straight face, are more acceptable forms of tracking software—known as “first-party” cookies. According to the tortured online advertiser explanation, “first-party” trackers are placed by websites on which you have affirmatively given permission to be tracked while you are on that site. These “we-have-your-permission-to-use” first-party cookies would increasingly become the foundation for advances in digital tracking and targeting. Please raise your hand if you believe you have informed Google or Amazon, to cite the two most egregious examples, that they can surveil what you do via these first-party cookies, including engaging in an analysis of your actions, background, interests and more. What the online ad business has developed behind its digital curtain—such as various ways to trigger your response, measure your emotions (link is external), knit together information on device (link is external) use, and employ machine learning (link is external) to predict your behaviors (just to name a few of the methods currently in use)—has played a fundamental role in personal data gathering. Yet these and other practices—which have an enormous impact on privacy, autonomy, fairness, and so many other aspects of our lives—will not be affected by the “death-of-the-cookie” transition currently underway. On the contrary, we believe that a case to be made that the opposite is true. Rather than strengthening data safeguards, we are seeing unaccountable platforms such as Google actually becoming more dominant, as so-called “privacy preserving (link is external)” systems actually enable enhanced data profiling. In a moment, we will briefly discuss some of the leading online marketing industry work underway to redefine privacy. But the motivation for this post is to sound the alarm that we should not—once again—allow powerful commercial interests to determine the evolving structure of our online lives. The digital data industry has no serious track record of protecting the public. Indeed, it was the failure of regulators to rein in this industry over the years that led to the current crisis. In the process, the growth of hate speech, the explosion of disinformation, and the highly concentrated control over online communications and commerce—to name only a few— now pose serious challenges to the fate of democracies worldwide. Google, Facebook and the others should never be relied on to defer their principal pursuit of monetization out of respect to any democratic ideal—let alone consumer protection and privacy. One clue to the likely end result of the current industry effort is to see how they frame it. It isn’t about democracy, the end of commercial surveillance, or strengthening human rights. It’s about how best to preserve what they call the “Open Internet.” (link is external)Some leading data marketers believe we have all consented to a trade-off, that in exchange for “free” content we’ve agreed to a pact enabling them to eavesdrop on everything we do—and then make all that information available to anyone who can pay for it—primarily advertisers. Despite its rhetoric about curbing tracking cookies, the online marketing business intends to continue to colonize our devices and monitor our online experiences. This debate, then, is really about who can decide—and under what terms—the fate of the Internet’s architecture, including how it operationalizes privacy—at least in the U.S. It illustrates questions that deserve a better answer than the “industry-knows-best” approach we have allowed for far. That’s why we call on the Biden Administration, the Federal Trade Commission (FTC) and the Congress to investigate these proposed new approaches for data use, and ensure that the result is truly privacy protective, supporting democratic governance and incorporating mechanisms of oversight and accountability. Here’s a brief review (link is external) of some of the key developments, which illustrate the digital “tug-of-war” ensuing over the several industry proposals involving cookies and tracking. In 2019, Google announced (link is external) that it would end the role of what’s known as “third-party cookies.” Google has created a “privacy sandbox (link is external)” where it has researched various methods it claims will protect privacy, especially for people who rely on its Chrome browser. It is exploring “ways in which a browser can group together people with similar browsing habits, so that ad tech companies can observe the habits of large groups instead of the activity of individuals. Ad targeting could then be partly based on what group the person falls into.” This is its “Federated Learning of Cohorts (FLoC) approach, where people are placed into “clusters” based on the use of “machine learning algorithms” that analyze the data generated from the sites a person visited and their content. Google says these clusters would “each represent thousands of people,” and that the “input features” used to generate the targeting algorithm, such as our “web history,” would be stored on our browsers. There would be other techniques deployed, to add “noise” to the data sets and engage in various “anonymization methods” so that the exposure of a person’s individual information is limited. Its TURTLEDOVE initiative is designed to enable more personalized targeting, where web browsers will be used to help ensure our data is available for the real-time auctions that sell us to advertisers. The theory is that by allowing the data to remain within our devices, as well using clusters of people for targeting, our privacy is protected. But the goal of the process— to have sufficient data and effective digital marketing techniques—is still at the heart of this process. Google recently (link is external) reported that “FLoC can provide an effective replacement signal for third-party cookies. Our tests of FLoC to reach in-market and affinity Google Audiences show that advertisers can expect to see at least 95% of the conversions per dollar spent when compared to cookie-based advertising.” Google’s 2019 announcement caused an uproar in the digital marketing business. It was also perceived (correctly, in my view) as a Google power grab. Google operates basically as a “Walled Garden (link is external)” and has so much data that it doesn’t really need third-party data cookies to hone in on its targets. The potential “death of the cookie” ignited a number of initiatives from the Interactive (link is external) Advertising Bureau, as well as competitors (link is external) and major advertisers, who feared that Google’s plan would undermine their lucrative business model. They include such groups as the Partnership for Addressable Media (PRAM), (link is external) whose 400 members include Mastercard, Comcast/NBCU, P&G, the Association of National Advertisers, IAB and other ad and data companies. PRAM issued a request (link is external) to review proposals (link is external) that would ensure the data marketing industry continues to thrive, but could be less reliant on third-party cookies. Leading online marketing company Trade Desk is playing a key role here. It submitted (link is external) its “United ID 2.0 (link is external),” plan to PRAM, saying that it “represents an alternative to third party cookies that improves consumer transparency, privacy and control, while preserving the value exchange of relevant advertising across channels and devices.” There are also a number of other ways now being offered that claim both to protect privacy yet take advantage of our identity (link is external), such as various collaborative (link is external) data-sharing efforts. The Internet standards groups Worldwide Web Consortium (W3C) has created (link is external) a sort of neutral meeting ground where the industry can discuss proposals and potentially seek some sort of unified approach. The rationale for the [get ready for this statement] “Improving Web Advertising Business Group goal is to provide monetization opportunities that support the open web while balancing the needs of publishers and the advertisers that fund them, even when their interests do not align, with improvements to protect people from the individual and societal impacts of tracking content consumption over time.” Its participants (link is external) are another “Who’s Who” in data-driven marketing, including Google, AT&T, Verizon, NYT, IAB, Apple, Group M, Axel Springer, Facebook, Amazon, Washington Post, Verizon, and Criteo. DuckDuckGo is also a member (and both Google and Facebook have multiple representatives in this group). The sole NGO listed as a member is the Center for Democracy and Technology. W3Cs ad business group has a number of documents (link is external) about the digital marketing business that illustrate why the issue of the future of privacy and data collection and targeting should be a public—and not just data industry—concern. In an explainer (link is external) on digital advertising, they make the paradigm so many are working to defend very clear: Marketing’s goal can be boiled down to the "5 Rights": Right Message to the Right Person at the Right Time in the Right Channel and for the Right Reason. Achieving this goal in the context of traditional marketing (print, live television, billboards, et al) is impossible. In digital realm, however, not only can marketers achieve this goal, they can prove it happened. This proof is what enables marketing activities to continue, and is important for modern marketers to justify their advertising dollars, which ultimately finance the publishers sponsoring the underlying content being monetized.” Nothing I’ve read says it better. Through a quarter century of work to perfect harvesting our identity for profit, the digital ad industry has created a formidable complex of data clouds (link is external), real-time ad auctions, cross-device tracking tools and advertising techniques (link is external) that further commodify our lives, shred our privacy, and transform the Internet into a hall of mirrors that can amplify our fears and splinter democratic norms. It’s people, of course, who decide how the Internet operates—especially those from companies such as Google, Facebook, Amazon, and those working for trade groups as the IAB. We must not let them decide how cookies may or may not be used or what new data standard should be adopted by the most powerful corporate interests on the planet to profit from our “identity.” It’s time for action by the FTC and Congress. Part 1. (1)For the uninitiated, TURTLEDOVE stands for “Two Uncorrelated Requests, Then Locally-Executed Decision On Victory”; FLEDGE is short for “First Locally-Executed Decision over Groups Experiment”; SPARROW is “Secure Private Advertising Remotely Run On Webserver”; and FLoC is “Federated Learning of Cohorts”). (2) In January 2021, the UK’s Competition and Markets Authority (CMA) opened up an investigation (link is external) into Google privacy sandbox and cookie plans.

The Center for Digital Democracy (CDD) announced today its opposition to the California Privacy Rights Act (CPRA), also known as Proposition 24 (link is external), which will appear on the November 2020 California general election ballot. CDD has concluded that Prop 24 does not sufficiently strengthen Californians’ privacy and may, in fact, set a new, low, and thus dangerous standard for privacy protection in the U.S. We need strong and bold privacy legislation, not weaker standards and tinkering at the margins. We need digital privacy safeguards that address the fundamental drivers of our eroding privacy, autonomy, and that redress the growing levels of racial and social inequity. We need rules that go to the heart of the data-driven business model and curtail the market incentives that have created the deplorable state of affairs we currently face. What we need are protections that significantly limit data uses that undermine our privacy, increase corporate manipulation and exploitation, and exacerbate racial and economic inequality. We need default privacy settings that limit the sharing and selling of personal information, and the use of data for targeted advertising, personalized content, and other manipulative practices. We need to ensure privacy for all and limit any pay-for-privacy schemes that entice the most vulnerable to give up their privacy. In other words, we need to limit harmful data-use practices by default, and place the interests of consumers above market imperatives by allowing only those data practices that are not harmful to individuals, groups, and society at large. Prop 24 does none of that. Specifically, Prop 24 continues on the path of a failed notice-and-choice regime, allowing the much more powerful companies to set unfair terms. Instead, privacy legislation should focus on strong default settings and data-use practices that are allowable (“permissible uses”) and prohibiting all others. These safeguards should be in place by default, rather than forcing consumers to opt out of invasive advertising. Prop 24, in contrast, does not provide effective data-use limitations; instead it continues to limit data sharing and selling via an opt-out, rather than declaring them to be impermissible uses, or at minimum requiring an opt-in for such practices. Even “sensitive data” under Prop 24 is protected only via a consumer-initiated opt-out, rather than prohibiting the use of sensitive personal data altogether. Equally concerning, Prop 24 would expand rather than limit pay-for-privacy schemes. Under the terms of Prop 24, corporations are still allowed to charge a premium (or eliminate a discount) in exchange for privacy. Consumers shouldn’t be charged higher prices or be discriminated against simply for exercising their privacy rights. This provision of Prop 24 is particularly objectionable, as it tends to harm vulnerable populations, people of color, and the elderly by creating privacy “haves” and “have-nots,” further entrenching other, existing inequities as companies would be able use personal data to profile, segment, and discriminate in a variety of areas. There are many other reasons that CDD objects to Prop 24, chief among them that this flawed measure - employs an outdated concept of “sensitive data” instead of focusing on sensitive data uses; - fails to rein in the growing power of data brokers that collect and analyze personal data from a variety of sources, including public data sets, for sale to marketers; - does not employ strong enough data minimization provisions to limit data collection, use and disclosure only to what is necessary to provide the service requested by the consumer; - undermines consumer efforts to seek enforcement of privacy rights by neglecting to provide full private right-of-action provisions; and - unnecessarily delays its protection of employee privacy.