Observatory
July 2026High-risk profiles in online financial services
Automated detection from Certificate Transparency, 30-day window. The figures describe indicators from automated analysis: they are not determinations. Definitive qualification rests with the competent authority.
Summary
Over the period the system produced 14,340 verdicts on newly issued domains with a financial profile. 5,119 show a high-risk profile (36% of the analysed sample), with recurring risk indicators for retail investors.
From sample to priority subset
- 14,340verdicts produced
- 5,119high-risk profile
- 1,669with targeting toward Italy
- 317strong targeting
Declared service type
| Service | Detections |
|---|---|
| CFD / forex broker | 1,882 |
| Crypto exchange | 917 |
| Trading firm | 761 |
| Subscription signal services | 516 |
| Investment advisory | 453 |
| Yield / staking | 233 |
CFD/forex broker remains the dominant type (~37% of high-risk profiles), followed by crypto. This is the recurring picture in retail-facing offerings.
Targeting toward Italian investors
| Targeting strength | Detections |
|---|---|
| Strong | 317 |
| Moderate | 419 |
| Weak | 933 |
| None | 3,450 |
~1,669 high-risk profiles show some targeting toward Italian investors, 317 of them strong (marked localisation, explicit appeals to the Italian public). This is the priority subset.
Recurring indicators
| Indicator | Frequency |
|---|---|
| No verifiable authorisation | 4,816 |
| AI-generated content | 2,802 |
| Guaranteed-return promises | 2,524 |
| Unverifiable testimonials | 2,447 |
| Deposit pressure | 1,746 |
| No identifiable legal entity | 1,718 |
| Unmatched licence claim | 847 |
| Artificial urgency | 761 |
Overall reading
The phenomenon appears industrialised (high share of AI-generated content, repeated patterns across type and indicators) and persistent in the trading/crypto segment. Explicit targeting toward Italy covers a minority of domains (~33% with non-zero targeting) but a numerically significant one: 317 strongly targeted in this period alone.
- AI-generated content in ~55% of high-risk profiles: points to industrialised site production, consistent with campaign-based operations.
- Guaranteed returns in ~49%: still the most widespread persuasive lever and a recurring marker in high-risk profiles.
Methodological note
Detection from Certificate Transparency (near-real-time), content extraction, classification with severity calibrated on popularity (CrUX) and registration data (RDAP). The figures are automated indicators subject to review; they are not determinations. They count verdicts produced in the period, not certificates observed: system throughput is a different funnel, reported on the page for authorities. Aggregate data available on request. Methodology →
Get the observatory when it's published
The picture for your brand
These are sector figures. The same analysis, on your brand, is free.