Platforms: it’s time to show your work
Transparency for online advertising is essential for holding platforms accountable
As a scientist, I’m judged not just by my analysis and findings: my work also must be reproducible by other researchers. They need access to the data and code that fuels my research to confirm my conclusions are valid. But when social media platforms report on their progress in reducing misleading or false advertising, they expect the public to accept their word on faith.
Given that recent Pew survey data show 64 percent of Americans believe that social media has a negative effect on the way things are going in the country, that’s not a defensible stance. With social trust attacks flourishing on platforms such as Facebook, YouTube, and Twitter, as well as history of violations of existing law, platforms should welcome independent scrutiny.
A pair of bills introduced recently in the U.S. Congress, if approved, would help increase accountability by requiring online platforms to “show their work.” These proposals would establish rules for transparency of digital advertising; one goes beyond advertising and sets broader regulation and transparency for algorithms as well. Most importantly, these proposals require online platforms to comply with higher standards, an enormous change from the status quo where transparency standards are set by the platforms themselves. Furthermore, those standards are voluntary, thus platforms get to pick and choose what information they make public, and to whom.
Promising proposals would set ad transparency standards
Both proposals establish rules to strengthen transparency for digital advertising, and one goes beyond and sets broader regulation and transparency for algorithms as well. There’s a lot to like in both bills, but there are some key differences between them.
On digital ads, the most important difference is this: The Social Media Disclosure and Transparency of Advertisements (DATA) Act, introduced by Rep. Lori Trahan, D., Mass., would make data about all digital ads, including targeting information, available but only to “academic researchers” and the Federal Trade Commission (FTC). This would keep this crucial information out of reach for journalists and advocates.
The Algorithmic Justice and Online Platform Transparency Act, introduced by Sen. Ed Markey, D., Mass, and Rep. Doris Matsui, D., Calif., also would require transparency of online advertising, including targeting information, but would release data to the public at large — not just academic researchers.
We believe that releasing data only to the FTC and academic researchers, as the the DATA Act specifies, is unnecessary and will hugely limit the usefulness of that bill. While researchers like myself could do good and useful work with this data, the fact is that journalists, as well as researchers not affiliated with a college or university, have been the ones, time and time again, to uncover public harms. Just a few examples:
- Craig Silverman, reporting for Buzzfeed, demonstrated Facebook neglected to include ads from political interest groups that meet its own criteria for “social issues, elections, or politics” in the Facebook Ad Library.
- An investigation by Jeremy Merrill for The Markup revealed that corporate interests such as ExxonMobil and Comcast often run paid ads that present quite different messages to conservatives and to liberals. but neither the platforms nor the advertisers disclose targeting information that reveals these strategies.
- Merrill also reported that Google was allowing advertisers to exclude nonbinary people from seeing ads, until his report brought this problem to the company’s attention, again via The Markup.
- Avaaz published this analysis showing Facebook repeatedly allowed misinformation to be run in political advertising in the leadup to the 2020 elections.
The Quest for Universal Ad Transparency
Aside from this difference on public availability of data, both legislative proposals would help toward our goal of universal ad transparency. As both bills recognize, we need transparency around all types of advertising.
Cybersecurity for Democracy’s work leading up to the 2020 elections demonstrated there is public interest in finding out who is paying for online political advertising and how they’re targeting ads to influence public opinion. However, these concerns go beyond advertising traditionally classified as “political,” which is complicated to define. Of particular worry are ads for housing, credit, and employment ads. The U.S. has laws on the books that require ads in these categories not be targeted in discriminatory ways. But right now the FTC lacks the information to actually verify that ads on digital platforms comply with these existing rules; both bills would fix that.
Of course, we need also to stay vigilant in protecting user privacy. Both of these proposals require this by setting out conditions that information be anonymized, reported in aggregate when appropriate, and other provisions to protect users. The purpose of both of these bills is to understand platform and advertiser behavior, not that of individual users, and protecting user privacy given those goals really isn’t that hard. Online platforms should not use privacy concerns as an excuse to not offer transparency when there are vetted methodologies to disclose information that increases public understanding.
Our data and modeling reveal how poorly Facebook has followed its own standards. And frankly, while we have often criticized Facebook’s policies and execution, it has done a better job than most platforms on providing data transparency. As the influence of online platforms grows, and examples of algorithmic spread of online mis- and disinformation continue to pile up, we must require platforms to show their work, so they can be held accountable by the public at large.
Cybersecurity for Democracy is a research-based effort to expose online threats to our social fabric — and recommend how to counter them. We are part of the Center for Cybersecurity at NYU.
Would you like more information on our work? Visit Cybersecurity for Democracy online and see how tools, data, investigations, and analysis are fueling efforts toward platform accountability. You can:
- Follow this Medium feed.
- Sign up for our newsletter at www.cybersecurityfordemocracy.org
- Find us on Twitter: @cyber4democracy and @lauraedelson2
