Documented institutional evidence
Different sectors, jurisdictions and causes. One recurring control question: where did the organisation assume integrity instead of proving it?
Selected cases
Each summary below is deliberately concise and paraphrased. Follow the primary-source link for the official finding, report or decision.
Official European sources documented discriminatory use of nationality-related data and severe failures of legal protection and redress in fraud-risk processes.
DQIntegrity lens: input selection and classification were not sufficiently lawful, explainable or challengeable before automated risk decisions affected families.
European Parliament study →The UK National Audit Office reported that automated transfer of Universal Credit National Insurance credits to HMRC had been suspended and that records needed correction.
DQIntegrity lens: a critical inter-system feed stopped operating as expected, creating incomplete contribution records unless the break was detected and repaired.
National Audit Office report →UK regulators concluded that operational-resilience and governance arrangements around the migration were inadequate, causing widespread disruption after launch.
DQIntegrity lens: this is a change-assurance example rather than proof that data was lost. It illustrates the need to test end-to-end processing, update paths, controls and operating readiness—not only a successful technical migration.
Official FCA source →The Horizon Inquiry and court judgments documented software defects, disputed branch-accounting data and profound failures in how system evidence was understood and used.
DQIntegrity lens: system outputs were treated as sufficiently reliable for consequential decisions without proportionate challenge of data integrity, known defects and evidential limitations.
Horizon Inquiry report →The FCA said Metro Bank failed to adequately monitor more than 60 million transactions with a value above £51 billion for money-laundering risk.
DQIntegrity lens: the existence of a monitoring platform does not prove that the full relevant transaction population is covered by effective controls.
Official FCA source →US authorities described serious anti-money-laundering failures in the Estonian branch and misleading statements about the effectiveness of controls.
DQIntegrity lens: monitoring, customer information, oversight and escalation did not create a trustworthy enterprise view of the relevant risk population.
Official SEC source →Sweden’s financial supervisor issued a warning and SEK 4 billion administrative fine after finding major deficiencies in anti-money-laundering work in Baltic subsidiaries.
DQIntegrity lens: local and group-level information, governance and control effectiveness did not provide a sufficiently reliable end-to-end risk picture.
Official Finansinspektionen source →ESMA’s review recorded that €1.9 billion of claimed escrow cash did not exist and examined supervisory and enforcement weaknesses surrounding the reporting.
DQIntegrity lens: high-consequence balances require independent source evidence, traceability and direct reconciliation—not reliance on internally supplied representations.
Official ESMA report →France’s Cour des comptes declined to certify the family-branch accounts for 2024 and reported €9.4 billion of errors that remained uncorrected nine months after payment.
DQIntegrity lens: detecting error is not enough. Sustainable control requires correction within the available window, clear ownership and evidence that the affected population has been addressed.
Official Cour des comptes source →
A sourced PDF with fuller case summaries, the seven-layer framework and questions to use when assessing a decision-critical data journey.
Different cases. The same need for proof.
A confidential diagnostic can identify which layer is producing the visible symptom—and which upstream assumptions remain untested.
