At Mozilla we believe that greater transparency into the online advertising ecosystem can empower individuals, safeguard advertisers’ interests, and address systemic harms. Lawmakers around the world are stepping up to help realize that vision, and in this post we’re weighing in with some preliminary reflections on a newly-proposed ad transparency bill in the United States Congress: the Social Media DATA Act.
The bill – put forward by Congresswoman Lori Trahan of Massachusetts – mandates that very large platforms create and maintain online ‘ad libraries’ that would be accessible to academic researchers. The bill also seeks to advance the policy discourse around transparency of platform systems beyond advertising (e.g. content moderation practices; recommender systems; etc), by directing the Federal Trade Commission to develop best-practice guidelines and policy recommendations on general data access.
We’re pleased to see that the bill has some many welcome features that mirror’s Mozilla’s public policy approach to ad transparency:
- Clarity: The bill spells out precisely what kind of data should be made available, and includes many overlaps with Mozilla’s best practice guidance for ad archive APIs. This approach provides clarity for companies that need to populate the ad archives, and a clear legal footing for researchers who wish to avail of those archives.
- Asymmetric rules: The ad transparency provisions would only be applicable to very large platforms with 100 million monthly active users. This narrow scoping ensures the measures only apply to the online services for whom they are most relevant and where the greatest public interest risks lie.
- A big picture approach: The bill recognizes that questions of transparency in the platform ecosystem go beyond simply advertising, but that more work is required to define what meaningful transparency regimes should look like for things like recommender systems and automated content moderation systems. It provides the basis for that work to ramp up.
Yet while this bill has many positives, it is not without its shortcomings. Specifically:
- Access: Only researchers with academic affiliations will be able to benefit from the transparency provisions. We believe that academic affiliation should not be the sole determinant of who gets to benefit from ad archive access. Data journalists, unaffiliated public interest researchers, and certain civil society organizations can also be crucial watchdogs.
- Influencer ads: This bill does not specifically address risks associated with some of the novel forms of paid online influence. For instance, our recent research into influencer political advertising on TikTok has underscored that this emergent phenomenon needs to be given consideration in ad transparency and accountability discussions.
- Privacy concerns: Under this bill, ad archives would include data related to the targeting and audience of specific advertisements. If targeting parameters for highly micro-targeted ads are disclosed, this data could be used to identify specific recipients and pose a significant data protection risk.
Fortunately, these shortcomings are not insurmountable, and we already have some ideas for how they could be addressed if and when the bill proceeds to mark-up. In that regard, we look forward to working with Congresswoman Trahan and the broader policy community to fine-tune the bill and improve it.
We’ve long-believed that transparency is a crucial prerequisite for accountability in the online ecosystem. This bill signals an encouraging advancement in the policy discourse.