Over the past several months I've been pulling public records on the wave of "age verification" bills moving through US state legislatures. IRS 990 filings, Senate lobbying disclosures, state ethics databases, campaign finance records, corporate registries, WHOIS lookups, Wayback Machine archives. What started as curiosity about who was pushing these bills turned into documenting a coordinated influence operation that, from a privacy standpoint, is building surveillance infrastructure at the operating system level while the company behind it faces zero new requirements for its own platforms.

The advocacy group that doesn't legally exist The Digital Childhood Alliance presents itself as a coalition of 50+ conservative child safety organizations (later inflated to 140+, though only six have ever been publicly named). It has been testifying in favor of these bills across states. Here is what public records show about its legal status:

DCA's domain was registered December 18, 2024 through GoDaddy with privacy protection and a four-year registration. The website was live and fully formed one day later: professional design, statistics, testimonials from Heritage Foundation and NCOSE staff, ASAA talking points already loaded. This is not a grassroots launch. This is a staging deployment of a pre-built site. 77 days later, Utah SB-142 became the first ASAA law signed in the country.

DCA processes donations through For Good (formerly Network for Good, EIN 68-0480736), which is a Donor Advised Fund. For Good explicitly states in its documentation that it serves "501(c)(3) nonprofit organizations." DCA claims 501(c)(4) status. DCA is classified as a "Project" (ID 258136) in the For Good system, not as a standalone nonprofit. I searched all 59,736 For Good grant recipients across five years, roughly $1.73 billion in disbursements. Zero grants to DCA, DCI, NCOSE, or any related entity. The donation page appears to be cosmetic.

Bloomberg reporters exposed Meta as a DCA funder in July 2025. The Deseret News detailed the arrangement in December 2025. No version of the website, across 100+ Wayback Machine snapshots, has ever disclosed funding sources. Every blog post and testimony targets Apple and Google. Meta is never mentioned or criticized.

Casey Stefanski, Executive Director, spent 10 years at NCOSE as Senior Director of Global Partnerships. Unusually, she never appears on any NCOSE 990 filing as an officer, key employee, or among the five highest-compensated staff. A senior director title at a $5.4M organization for a decade with no 990 appearance suggests either below-threshold compensation, an inflated title, or something else about the arrangement.

NCOSE's own 501(c)(4) structure turns out to be complicated. Tracing Schedule R filings across four years reveals that NCOSE created "NCOSE Action" (EIN 86-2458921) as a c4 in 2021, reclassified it from c4 to c3 in 2022, then created an entirely new c4 called "Institute for Public Policy" (EIN 88-1180705) in 2023 with the same address and the same principal officer (Marcel van der Watt). By 2024 the original entity had disappeared from Schedule R entirely.

$70M+ in super PACs, deliberately fragmented Meta poured over $70 million into state-level super PACs and structured every one to avoid the FEC's centralized, searchable database:

If you maintain software that could be classified as an "operating system provider" under these definitions, start Full dataset, OSINT tasklist, and all processed findings are published with sources embedded in each file: github.com/upper-up/meta-lobbying-and-other-findings

  - a 16 hour old HN account sharing
  - an 18 hour old github account with
  - a 17 hour old repo