9 September 2026. Bounded acquisition closed. A named Danish grant is established, with a programme purpose more specific than general organizational support. Later recipient reporting also changes the temporal picture: foundations dominate the2025funding graphic, whereas major donors dominate2024. These sources establish published purposes, resource relationships and internal foundation authority, not a command to undertake the February2025Senate intervention.
KR Foundation's2023annual-report grant table, printed46/PDF24 of the retained double-page republication, lists Check My Ads, US, DKK1,991,232 granted in2023, start1May2023, end1May2025. Its project is titled 'Dismantle the disinformation economy: cut climate change and election disinformation off at the source.' Root visually inspected the row, headers and amount. The old [donor-hosted report address](http[local research file] remains indexed with the same table but ordinaryGET returned404. Root obtained the report from an [Americans for Public Trust republication](http[local research file] this is a donor-authored document in third-party custody, not an endorsement of the host's framing or a fresh donor original. SHA301137e051f62930e4be97813bdd0996f81a5fd7d4778007bd329128855613d5.
The live donor website links a 2023grant list.
The records identify actual Danish support consistent with CMA's own declaration. They do not establish that KR was its only Danish supporter or permit allocating the whole grant to calendar2024or the Senate campaign. DKK is retained without a speculative exchange-rate conversion. The2025award begins after the Februaryletters; the earlier award term overlaps them, but its named purpose concerns climate and election disinformation, not CSAM.
The same2023report, printed16–17/PDF9, explicitly arranges supported organizations around regulatory change, pressure on fossil-fuel service providers, and capacity building. Root viewed the actual diagram after extraction produced repeated hidden text. It places CMA, Center for Countering Digital Hate and Stop Funding Heat in the same category concerning accountability of enablers. It places Friends of the Earth/CAAD in a separate capacity-building category. This is the funder's published programme architecture. It proves neither that every listed organization collaborated on one campaign nor that any one directed another. The stated2023geographic priorities include Europe, USA and Australia. Those programme statements, together with the named grant, are stronger evidence of intentional funding around an agenda than mere overlap in a donor list.
The current [How we work page](http[local research file] substantive primary reader, says applications are invitation-only and grants board-approved. OrdinaryGET returned455; that response is not the substantive page. The [currently linked statutes translation](http[local research file] original captured and all five pages read, supplies more exact authority. Section5 assigns distribution, eligibility, scale and timing to the board and permits defined delegation to the chairmanship/executive director. Section7 entitles VILLUM FONDEN to appoint up to two of five-to-seven board members; remaining seats are filled by board election. Election of those remaining members requires a majority with no more than one dissent. This means two dissenting eligible directors can block such an election; it does not prove that the founder-appointed directors coordinated votes or can individually dictate grants. Ordinary decisions use simple majority and a chair casting vote. Section8 specifies joint signatures or delegated power to bind the foundation. This is internal KR authority, not authority over CMA's publications.
Sections1and4 identify VILLUM FONDEN and descendants of Villum Kann Rasmussen as founders and a DKK10million initial cash capital, with no return of that initial capital to founders. No present grant is traced to a particular upstream transfer, and no wider family, company or state financing claim is made. The original first-board names in the statutes are historical, not today's board. Current bios were read only as context and are not used to retroactively assign the2023grant vote.
No executed CMA award agreement, approved proposal, reporting template, tranche condition, clawback, editorial approval, target list or evaluation was found in the inspected grant lists, reports, statutory text and two focused searches. Those instruments remain unacquired; invitation-only selection is an actual programme-entry gate, not proof of later editorial veto. The linked2023financial report was located and the opening reader through the auditor's introductory scope was inspected; its financial totals and grant accounting have not been adopted here.
The [CMA2025annual report](http[local research file] original SHA570059f9a0e37ddbf5993033765238a4717d089f39b81db3c02ba8632e4193a2, PDF35 visually inspected, explicitly reports foundation grants82.8%, major donors12.3%, individual donors4.9%, four new institutional donors and three multi-year renewals. It also adds UK and Canadian foundation support alongside USA and Denmark. This supersedes any implication that2024's major-donor majority describes the later mix. Absolute income, category definitions and accounting basis still are not given on that page. The shift may combine new funding, reduced large gifts, recognition timing or recategorization; the two graphics alone do not resolve the cause or identify every new funder.
Scoped report text read atPDF26,29,32and35 adds direct institutional leads: CMA claims a TAG complaint, support for industry KYC and MRC changes; it describes an Ofcom Online Safety Act additional-safety submission seeking tighter monetization controls. Its policy platform names KYC due diligence and mandatory supply-chain transparency alongside privacy-respecting advertising. These are the organization's claims and proposals, not observed enforcement or an established duty for all internet users to identify themselves. The actual Ofcom submission and its proposed subjects, triggers and safeguards would decide more than inference from the initials KYC. The rest of the37-page report has been captured, not comprehensively reviewed; incidental excerpts are not a whole-report audit.
Secondary search also surfaced a2025Oak Foundation grant claim; it remains unverified discovery for a later named-funder reconciliation, not an accepted edge here. No unsolicited expansion into donor genealogy occurred.
Per-file captureJSON records status, URLs, bytes and hashes. The original2023list SHA4cad141c609ac4de80c101d1bcce82f7fc291e62c0470ef60e77117477404742; original2025list SHA40704711dab007be9129e88e933134a8fc15bf4532cde43800e2c1a968a0e65d; originalstatutes SHA14806448fb046190304f56f5471239aeb3405645d176910d9396040bec8a8fa7. The failed oldreport request is a404HTML response, originally named.pdf by the capture script; its manifest explicitly saysoriginal_pdf=false. It was never parsed as aPDF or treated as missing evidence of suppression. PNGs of the table, programme diagram,2025list and CMA2025chart were actually viewed. Sources denied or missing were closed; no challenges, accounts, visible UI, outreach or ZIP work occurred.