- Home
- Family Scam Pulse
- Quarterly Reports
- Q2 2026
Family Scam Pulse · Quarterly Report
AI Scam Report — Q2 2026
Reporting period: April 1, 2026 – June 30, 2026 · Data last updated August 26, 2026
Key findings
- 5 verified AI impersonation incidents were added to the Family Scam Pulse database in Q2 2026, spanning 2 countries.
- $5,400 in disclosed losses across the 1 incident where the source reported an explicit US-dollar amount. This is a floor, not an estimate — 4 incidents had losses that were undisclosed or not concretely quantified.
- Voice Cloning was the most common attack type (5 of 5 incidents).
- The most frequently targeted group was family (5 of 5 incidents).
- 1 incident was prevented before money changed hands — verification works when it happens.
Charts
All charts are plain counts of verified incidents. Each is available as a standalone SVG you can embed with attribution baked in.
Embed code
<img src="https://trustonion.io/family-scam-pulse/reports/2026-q2/charts/victim-types.svg" alt="Horizontal bar chart of verified AI scam incidents added in Q2 2026 by victim type: Family 5." width="640" /> <p>Source: <a href="https://trustonion.io/family-scam-pulse/reports/2026-q2">Trust Onion Family Scam Pulse, Q2 2026 report</a></p>
Embed code
<img src="https://trustonion.io/family-scam-pulse/reports/2026-q2/charts/outcomes.svg" alt="Horizontal bar chart of verified AI scam incidents added in Q2 2026 by outcome: Successful Fraud 3, Ongoing 1, Prevented 1." width="640" /> <p>Source: <a href="https://trustonion.io/family-scam-pulse/reports/2026-q2">Trust Onion Family Scam Pulse, Q2 2026 report</a></p>
Embed code
<img src="https://trustonion.io/family-scam-pulse/reports/2026-q2/charts/countries.svg" alt="Horizontal bar chart of verified AI scam incidents added in Q2 2026 by country: United States 4, Ireland 1." width="640" /> <p>Source: <a href="https://trustonion.io/family-scam-pulse/reports/2026-q2">Trust Onion Family Scam Pulse, Q2 2026 report</a></p>
Incidents in this report
Bay Area, California, United States · voice cloning · successful fraud · $5,400
Original reporting: ABC7 News (KGO San Francisco)
Ireland · voice cloning · ongoing · undisclosed
Original reporting: TechBuzz Ireland
Missouri, United States · voice cloning · successful fraud · Thousands of dollars (exact amount undisclosed)
Original reporting: WDBJ7
Vancouver, Washington, United States · voice cloning · prevented · undisclosed
Original reporting: Google Alerts
United States · voice cloning · successful fraud · undisclosed
Original reporting: CBS News
Source data
The exact records behind this report are available as CSV for Q2 2026, and the full database as a complete CSV. Every record links to our written case analysis and the original news reporting.
Methodology
- The reporting period covers incidents verified and added to the database during Q2 2026 (April 1, 2026 – June 30, 2026). Source-reported incident dates are shown per record and may precede the period, since scams often surface in the press weeks after they happen.
- Every incident passed the Pulse inclusion gate: reported by an identifiable news outlet or official body, involving AI-enabled impersonation, with a verified attack type, outcome, and working source link.
- Loss totals sum only explicitly-USD disclosed amounts. Amounts in other or unconfirmed currencies are shown verbatim and never converted or summed. Vague figures ("thousands of dollars") are excluded from totals.
- All statistics are direct counts or sums over the published records — nothing is modeled, extrapolated, or estimated.
Full inclusion, verification, and corrections rules: Family Scam Pulse methodology.
Limitations
- This is a sample of press-reported incidents, not a census. Most AI impersonation scams are never reported to media, so counts here dramatically understate real prevalence. Do not read quarter-over-quarter counts as market-wide trends.
- With 5 incidents this quarter, the sample is small; category breakdowns describe this dataset, not the population of all scams.
- Loss totals are a floor: 4 incidents had undisclosed losses and 0 had losses without a concrete figure.
- Geographic coverage reflects English-language press monitoring and skews toward the United States and other English-speaking countries.
How to cite this report
Free to cite and reuse with attribution (CC BY 4.0). Suggested citation:
Trust Onion, "AI Scam Report — Q2 2026," Family Scam Pulse, https://trustonion.io/family-scam-pulse/reports/2026-q2
Please also credit the original news source linked from each incident record. For interviews or rapid commentary on breaking AI scam stories, see the press page or email press@trustonion.io — we respond to journalist deadlines within one business day.