TruaceTracing the truth around AIWednesday, August 26, 2026
TRV-2026-0084Certified recordJournalism

Investment fraud in UK soared to more than £220m lost last year, trade body says

Increasingly elaborate investment scams involving gold, cryptocurrencies and wine have soared in the past year with more than £220m lost to the fraud, according to a report. UK banks reported almost 15,000 investment scams in 2025 as criminals use artificial intelligence to dupe people out of their money. About £221.5m was lost to scams in which people were persuaded to move their money to a fake investment or a fictitious fund, a rise of 40% from the year before, according to the report from UK Finance. This parti

Crime · P Space — documented harm · certified 2026-07-12 · v4 · article view · machine-readable

Current reading — problem

Criminals using artificial intelligence to run investment scams persuaded UK victims to move money into fake funds, causing about £221.5m in losses in 2025

What this doesn’t fix

No countervailing benefit is described in the supplied text

Evidence

Reader signal

How should this claim be treated?

Cite this record

Truvace Impact Record TRV-2026-0084, v4: “Investment fraud in UK soared to more than £220m lost last year, trade body says.” Truvace, 2026-07-13. /record/TRV-2026-0084 (accessed at citation time). sha256 60bfad7e432183c0

Calibration history

Every change to this record since certification, in the open.

  1. Revisedv460bfad7e4321

    Model backfill: grounded claim, summary, sector, and trace validation

  2. Certifiedv3e8c4c48dcea8

    Restored after model confidence scale normalization

  3. Retractedv232c2439249c8

    Model backfill: source did not support a publishable AI-impact claim

  4. Certifiedv1853f7b29e6fe

    Certified into the record

Verify this record
How to verify without trusting this page

Fetch the canonical text of any version from /api/record/TRV-2026-0084 and hash it yourself — for example shasum -a 256 on the saved canonical field. The result must equal content_hash, and each version’s text ends with prev:followed by the prior version’s hash (version 1 chains to 64 zeros). If a single character of any version had been altered since certification, the chain would not reproduce.