Thread by @doctorow

I've been working with @EFF for 20 years (!) now, and that association continues to pay dividends. EFF basically invented the idea of promoting tech policy positions that were informed by deep expertise in technology, law and human rights principles. 1/

If you'd like an unrolled version of this thread to read or share, here's a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog:

pluralistic.net/2022/03/22/myo… 2/

This three-legged stool produces some remarkably sturdy proposals and policies - proposals that are legally sound, technologically achievable, and that advance important human rights causes in the digital realm. 3/

The latest of these is EFF's policy on how and why to ban surveillance advertising, entitled "Ban Online Behavioral Advertising," written by my colleagues @Adam_D_Schwartz and Bennett Cyphers:

eff.org/deeplinks/2022… 4/

The article begins by laying out the mechanisms that make behavioral ads - that is, ads that "[target] us based on our online behavior" - viable: 5/

I. Tracking: When our tech "quietly collects information" about who we are and what we do, e.g. "app interactions and browsing history," and share that info with "ad tech companies and data brokers" 6/

II. Profiling: Linking that data to information that ad tech/data broker companies have on us already, then drawing inferences about our gender, age, what we like, and what we might buy, attend or vote for 7/

III. Targeting: Using profiles to decide which tailored messages are delivered to "particular people, types of people, or groups" on "websites, apps, TVs, and social media" 8/

Noting that "what’s good for an advertiser is actively harmful for their targets," the paper explains how these three processes harm users: 9/

* Behavioral data isn't just used for ads, it is "shared with or sold to hedge funds, law enforcement agencies, and military intelligence"

* The data is "accessed and exploited by people inside the company for personal ends" 10/

* Surveillance advertising has "warped the development of technology so that our devices spy on us by default" ("advertising IDs" in mobile phones, the continued existence of third-party cookies despite serious security issues) 11/

* Surveillance ads "[supercharge] the efforts of fraudulent, exploitive, and misleading advertisers" 12/

* They allow "peddlers of shady products and services to reach exactly the people who, based on their online behavior, the peddlers believe are most likely to be vulnerable to their messaging" 13/

* Companies like Facebook "run automatic experiments to identify exactly which kinds of people are most susceptible to a particular message...targeting credulous users with deceptive ads for modern-day snake oil" 14/

* Some examples: targeting subprime finance products to the "financially vulnerable" and marketing "investment scams to thousands of seniors"

* Surveillance ads have discriminatory impacts, allowing advertisers to target based on "gender, age, race, religion" 15/

* Even when these categories aren't explicitly targeted, ad profiles allow for targeting based on proxies: "location, purchase history, credit status, and income"

* All of this enables "turnkey housing discrimination and racist voter suppression" 16/

Given all of those information security and human rights problems, the authors turn to how to craft legislation that both passes constitutional muster and will be difficult for ad-tech companies to subvert. 17/

They home in on "the personal data most central to targeted ads: our online behavior." That includes "the web searches we conduct, the web pages we visit, the mobile apps we use, the digital content we view or create, and the hour we go online." 18/

But also the "offline" info our devices collect "such as our phones using GPS to track our geolocation or fitness trackers monitoring our health." 19/

Congress should "ban any entity that delivers online ads from doing so by targeting users based on their online behavior...whether or not an ad is targeted to a traditional personal identifier, like a name or email address." 20/

(What's an ad? "Paid content that concerns the economic interests of the speaker and audience.") 21/

It's not enough to simply address the visible parts of the ad tech industry. Legislators should also regulate data brokers, who "sell lists to advertisers" even if they don't deliver ads: 22/

"Legislators should ban an ad deliverer from using a list created by another entity, if the deliverer knows it is based on users’ online behavior, or would have known but for reckless disregard of known facts." 23/

And brokers should be banned from "disclosing a list of users that is based on online behavior, if the data broker knows it will be used to deliver ads, or would have known but for reckless disregard of known facts." 24/

The exceptions should be related to what a user's "rough locations" and what they're is doing *right now*, in order to enable low-surveillance, privacy-respecting "contextual ads" (ads based on what you're reading/doing, not who you are).

pluralistic.net/2020/08/05/beh… 25/

These proposals are part of a wider package of reforms that EFF backs, especially the idea of a US federal privacy law with a private right of action that would allow users (or orgs that take up users' cases) to sue companies that violate it:

eff.org/deeplinks/2019… 26/

Though this proposal is focused on users' human rights, it's also a win for publishers. 27/

From the media's perspective, banning surveillance ads and promoting contextual ads is far superior to the "link taxes" that have been tried in #Australia and the #EU and may be coming to #Brazil.

eff.org/deeplinks/2022… 28/

These just set the stage for media monopolies to collude with tech monopolies to freeze out both independent reporting and independent tech - and they also have grave consequences for free expression, limiting our ability to discuss the news and its coverage. 29/

Image:

EFF
eff.org/files/banner_l…

CC BY 3.0
creativecommons.org/licenses/by/3.… 30/