Study: 86K+ Facebook pages ran at least one political ad that was not properly disclosed, and ~20K ads had been purchased by “likely inauthentic communities”
https://www.washingtonpost.com/ ... Rasmus Kleis Nielsen / @rasmus_kleis : New research “point to myriad opportunities malicious actors may have had to exploit the platform's powerful targeting tools while hiding their tracks, misleading users and evading Facebook's enforcement” @TonyRomm and @isaacstanbecker report https://www.washingtonpost.com/ ... See also Mediagazer
Context & Ripple Effects
The disclosure-gap story has been building since 2018, when ProPublica found 12 Facebook campaigns masking who paid for political issue ads and VICE reporters got Facebook to approve fake ads purchased in the names of all 100 US senators. What was then framed as isolated enforcement misses now looks systemic: this study counts 86,000+ pages running at least one improperly disclosed political ad.
The scale matters because it reframes the problem from bad actors slipping through to a structural weakness in self-serve political ad buying — echoing the earlier finding that 122 of 228 groups running divisive 2016 election ads were suspicious, some tied to the IRA. Four years later, ProPublica documented eight deceptive operations on Meta's platforms still collecting user data through 160K+ ads.
First-order effects
- Facebook's political-ad transparency tools are directly implicated: researchers could quantify undisclosed spending across tens of thousands of pages, handing regulators and journalists a concrete enforcement case rather than anecdotes.
- Legitimate political advertisers now operate inside a disclosure regime shown to be porous, meaning their own ads sit alongside unattributed ones competing for the same targeted audiences.
Second-order effects
- Enforcement economics invert: with 86K+ offending pages versus the handful of campaigns in earlier investigations, manual review cannot keep pace, pushing Facebook toward automated detection whose errors will hit compliant advertisers too.
- The recurring pattern of fake-ad approvals and masked sponsorship invites legislative fixes that would impose disclosure requirements on all platforms selling political ads, not just Facebook.
Third-order effects
- If self-serve targeting keeps outpacing verification, political advertising on social platforms drifts toward an environment where attribution is optional — the same structure the 2024 ProPublica operations exploited for data collection.
- Sustained findings like these make paid political reach on social platforms a regulatory target comparable to broadcast rules, shifting compliance costs onto platforms and reshaping how campaigns buy digital persuasion.
The trend: Social platforms' self-serve political ad machinery keeps outrunning disclosure enforcement, turning ad transparency from a policy feature into a persistent cat-and-mouse between researchers, malicious buyers, and regulators.