NumbersOnline
Data report · September 2026

Phone spam trends — September 2026

Reported phone activity in September 2026, aggregated from regulator complaint feeds, our honeypot fleet, crawled complaint sites, and community reports. Figures were snapshotted at publish time and are never restated. Everything here is a supplementary signal — advisory, not a verdict.

← All reportsAugust 2026
121.9K
Numbers with observed activity
113.3K
Newly observed this month
350K
Reports across sources
63
Countries with activity

Key findings

  • In September 2026, 121,948 phone numbers had reported activity across 63 countries, and 113,281 of them were observed for the first time that month.
  • The numbers active in September 2026 carried 349,990 reports across all sources — cumulative totals as of the snapshot's publish date, not in-month counts.
  • United States had the most numbers with reported activity in September 2026 (93,384), followed by Poland (20,825) and Denmark (3,199).
  • Among the countries listed in both the September 2026 and August 2026 snapshots, United Kingdom saw the largest increase in active numbers (+364) and United States the largest decrease (−22,042).
  • The most common report category in September 2026 was Robocall, the dominant category for 64,846 active numbers, ahead of Reported activity (26,024).
  • Numbers newly observed in September 2026 clustered most in the +1 833 range (United States, 2,212 numbers).

Country movement

vs August 2026
93.4K active
88.8K new
−22,042
Robocall
Poland
20.8K active
20.5K new
+99
Reported activity
Denmark
3,199 active
212 new
+229
Reported activity
1,682 active
1,526 new
+364
Phishing / smishing
Portugal
647 active
578 new
+133
Reported activity
Germany
392 active
327 new
+188
Reported activity
Italy
278 active
243 new
+82
Reported activity
Latvia
212 active
147 new
+18
Reported activity
Estonia
170 active
99 new
+23
Reported activity
Lithuania
167 active
82 new
−46
Phishing / smishing
Slovenia
133 active
99 new
+44
Reported activity
Croatia
105 active
85 new
+14
Reported activity
Bulgaria
99 active
77 new
+32
Reported activity
Romania
98 active
55 new
−17
Phishing / smishing
France
79 active
76 new
+5
Reported activity
Spain
61 active
59 new
+15
Reported activity
Belgium
39 active
27 new
+39
Reported activity
Netherlands
38 active
24 new
−7
Reported activity
Slovakia
34 active
33 new
+34
Reported activity
Sweden
33 active
16 new
+33
Reported activity
Active = numbers whose observed-activity window overlaps September 2026. Movement compares publish-time snapshots, never live data. Country pages: browse the corpus.

What got reported

Robocall−12364,846
Reported activity+50426,024
Telemarketing−2,81925,488
Phishing / smishing−18,5322,540
Debt collection+782,065
Auto warranty−156546
Harassment+49215
Silent / hang-up+58160
Survey±041
Ping call−1323
Categories are low-confidence labels describing what reporters allege — a supplementary signal, not established fact.

New number-range patterns

+1 8332,212+1 8772,037+1 8881,700+1 8551,357+1 8661,354+1 8441,016+1 502978+1 813791+1 804719+1 941599+1 904590+1 903543
Ranges where numbers newly observed in September 2026 cluster (three or more). A range is never bad by itself — plenty of legitimate numbers share it.
Snapshot published October 3, 2026. Figures are frozen as of that date.

Data downloads

This month’s frozen figures as CSV — aggregates and prefixes only, no individual phone numbers. Cite as “Numbers Online, Phone spam trends — September 2026”.

Methodology

These figures aggregate our phone-intel corpus: regulator complaint feeds (among them the US FTC Do Not Call and FCC consumer complaint data and several European regulators), our own honeypot number fleet, independent complaint sites we crawl, and reports filed by the Numbers Online community. Only publicly displayable sources are counted, and numbers verified by their owner as personal are excluded throughout.

A number counts as active in a month when its observed-activity window overlaps that calendar month (UTC); newly observed means it first entered our corpus during the month. Activity dates reflect when a source last observed a number — never when we crawled it. Report totals are the cumulative counts our sources carried at publish time, not in-month deltas.

Each month’s figures are snapshotted when the report is published and never restated afterwards, so historical reports stay stable even as the live corpus grows. Month-over-month movement compares one publish-time snapshot with the previous one — never a frozen snapshot with live data.

Everything here is a supplementary signal — advisory, not a verdict on any number or its owner. Category labels are low-confidence and describe what reporters allege, not established fact. A number range is never bad by itself: plenty of legitimate numbers share every prefix shown. No individual phone numbers appear in these reports — aggregates and prefixes only.

Got a call worth reporting? Report it · think something is listed in error? Request removal.