Industry research puts US identity fraud losses at $27.3 billion for 2025, and combined identity fraud and scam losses at approximately $38 billion across around 36 million victims, with about 18 million people specifically victims of identity fraud.
This Is The Number The Corpus Has Been Missing
Three hundred files in this database describe organisations losing data. Almost none of them can say what happened to the people in it. The affected counts are precise — 6.5 million at Co-op, 5,556,702 at Yale New Haven, 62.4 million claimed at PowerSchool — and each file ends with credit monitoring and an open question.
This is the other end. Not what was taken, but what was lost, by whom, in a year.
And It Does Not Connect To The Other End
The corpus cannot join these two datasets, and neither can anybody else. There is no mechanism that traces a fraud in November back to a breach in March. The victim does not know. The bank does not know. The breached organisation certainly does not.
So the honest position is that this file and the incident files describe the same phenomenon from two sides with no bridge between them, which is the argument set out at 25-1219.
The Composition Matters More Than The Total
$38 billion combines two different things. Identity fraud is somebody using your details without your involvement. Scams are somebody persuading you to act.
They have different causes, different remedies and different relationships to the incidents in this database. Reporting them as one figure is convenient and analytically unhelpful, and the split — roughly $27.3 billion against the remainder — is the more useful part of the finding.
Graded medium: these are modelled estimates from consumer survey research, not counts. The FTC’s directly reported figure is substantially lower, which is the subject of 25-0509.
Built on published industry research, listed below, derived from consumer surveys and modelled to national totals. It is not a count of confirmed incidents. Corrections: corrections@forensicpost.com.