Only 50.4% of the 355,408 UK registered charities cross-reference with other data sources. The remaining 176,000-plus organisations exist, in data terms, almost entirely in isolation — creating a substantial blind spot for grant-makers, compliance teams, and due-diligence professionals who rely on conventional company intelligence.
BORSCH.AI processed the full Charity Commission register alongside 145,736 annual return records to produce this analysis. Here is what the numbers reveal about transparency, compliance gaps, and the hidden complexity of the UK’s third sector.
The Scale of the UK Charity Sector in Data
The Charity Commission API delivered 355,408 charity records — but when cross-referenced against BORSCH.AI’s wider dataset of 16.1 million UK entities, only 179,350 produced a meaningful match. That 50.4% match rate is the most telling figure in this entire dataset.
| Data Source | Records Fetched | Records Matched | Match Rate |
|---|---|---|---|
| Charity Commission Register | 355,408 | 179,350 | 50.4% |
| Charity Annual Returns (Part A) | 145,736 | 56,863 | 39.0% |
The gap is not a technical anomaly. It reflects the structural reality of the charity sector: roughly half of registered charities are unincorporated associations or charitable trusts — legal forms that do not require Companies House registration. They have no PSC filings, no filed accounts under the Companies Act, no director appointments on the public record. For anyone performing due diligence using company data alone, these organisations are functionally invisible.
The annual return picture is even starker. Of 145,736 annual return records fetched, only 56,863 — 39.0% — matched to an entity in BORSCH.AI’s broader database. That means the majority of charities submitting financial data to the Charity Commission have zero presence in the commercial intelligence ecosystem most compliance teams actually use.
Signal Density: What Charities Tell Us Compared to Commercial Entities
BORSCH.AI’s platform aggregates 50,024,066 signals across 53 data sources. The two charity-specific sources contribute a combined 400,498 signals — 272,572 from the Charity Commission register and 127,926 from annual returns.
| Signal Source | Signals Generated | % of Platform Total |
|---|---|---|
| Charity Commission | 272,572 | 0.54% |
| Charity Annual Returns | 127,926 | 0.26% |
| Combined Charity Sources | 400,498 | 0.80% |
| Companies House Officers | 22,449,602 | 44.9% |
| Companies House PSC | 11,897,682 | 23.8% |
A commercial company with active directors, filed accounts, and any mortgage or charge activity might generate dozens of signals across multiple sources. A registered charity generating signals from just two sources — and only when it clears a match threshold — produces a fraction of that intelligence trail.
This signal sparsity has direct implications for risk assessment. Conventional entity screening tools calibrated for commercial entities will systematically under-score charity-sector risk, not because the risk isn’t there, but because the data infrastructure to surface it is thinner.
The Annual Return Gap: A Compliance Red Flag
The Charity Commission’s annual return data had a sync failure at the time of this analysis — the source returned an error during its most recent processing run. Despite this, 145,736 records were successfully fetched in prior cycles. The fact that only 56,863 of those records matched company-level profiles confirms what charity governance specialists already suspect: the majority of charities required to submit annual returns are doing so in a data silo, with no corresponding footprint in the commercial registry system.
For compliance officers and grant-making bodies, this creates a specific problem:
- A charity can submit annual return data showing income, expenditure, and activities — but without a Companies House number, cross-checking that financial data against officer appointments, charges, or PSC filings is not straightforward
- Unincorporated charities are governed primarily by their trustees, whose personal details do not appear in the same public registers as company directors
- The 39.0% match rate on annual returns means over 88,000 annual return records exist in a regulatory data layer that most due-diligence workflows never reach
What Due Diligence Teams Are Missing
The governance category accounts for the largest share of all signals on BORSCH.AI’s platform — 32,809,178 signals, or 65.6% of the total. But charity governance data contributes a disproportionately small share of that figure, given that 355,000 organisations are involved.
For professionals assessing charities — whether as grant recipients, public sector contractors, or beneficiary organisations — the practical gaps are:
Grant-makers and foundations: The 360Giving Grants Registry (138,039 signals) and Contracts Finder (202,987 signals) together provide a partial picture of charities receiving public and philanthropic money. But without connecting those grants to the charity’s annual return data and regulatory status, funders are working with incomplete information.
Public sector procurement teams: Charities are active participants in public contracts. Contracts Finder holds 202,987 matched records across all entity types. Identifying which of those belong to charities — and then assessing those charities’ financial health through annual return data — requires exactly the kind of cross-source matching that BORSCH.AI performs, but that manual research cannot replicate at scale.
AML and fraud compliance: The ICO Data Protection Register (600,000 signals) and HMRC’s Anti-Money Laundering supervised business register (26,538 signals) provide compliance touchpoints for commercial entities. Charities face the same obligations but their compliance footprint is fragmented across the Charity Commission’s own registers rather than consolidated in commercial data sources.
The 50% That Does Cross-Reference — And Why It Matters
The 179,350 charities that did match BORSCH.AI’s broader dataset are, almost by definition, the more institutionally substantial ones: charitable companies limited by guarantee (registered at Companies House), charitable incorporated organisations (CIOs), and larger bodies that have acquired a Companies House number through subsidiary structures or trading arms.
These are also the charities most likely to hold public contracts, receive large grants, employ staff at scale, and carry reputational or financial risk for funders and partners. The fact that BORSCH.AI can surface governance, financial, and compliance signals for this 50.4% cohort — drawing on Companies House officers data, XBRL accounts, PSC filings, and the charity-specific sources simultaneously — represents a meaningfully different analytical capability compared to querying either register in isolation.
For the other 49.6%, the Charity Commission and annual return data are often the only structured data sources available. That makes the quality, completeness, and cross-referencing of those sources more critical, not less.
What This Means for Your Organisation
If you fund charities: Your due-diligence workflow should explicitly distinguish between incorporated and unincorporated charities. The data infrastructure for each is different, and the risk profile is not the same.
If you screen charities for AML or fraud: A charity with no Companies House presence is not inherently suspicious — but it does require a different screening protocol. Relying on company-centric tools will produce false confidence.
If you are a charity trustee or finance officer: The annual return sync failure observed in this dataset is a reminder that public data pipelines break. Ensuring your charity’s information is accurate and current across both the Charity Commission register and any relevant Companies House filings reduces the risk of being misidentified or overlooked in third-party assessments.
The UK charity sector’s 355,000-plus organisations represent an enormous component of civil society — but their data footprint is fragmented in ways that create genuine intelligence gaps. Closing those gaps requires connecting sources that were never designed to talk to each other.
Explore BORSCH.AI’s full charity intelligence capability, including cross-source matching across 53 data sources, at borsch.ai.
Disclaimers
Disclaimer: This article was generated with AI assistance using data from Borsch.AI’s aggregation of 53 UK government sources. While all statistics are derived from real data, analysis and interpretation are AI-generated and should be independently verified.
Disclaimer: Data presented reflects information available at the time of publication and may not reflect the most current state. Source data is aggregated from public government registers which may contain delays, errors, or omissions.
Comments