DEV Community

IP99
IP99

Posted on

This Week's Proxy Exits: 10.7M New Addresses, Still Almost All Off Datacenter Space

Datacenter blocklists still look like the default proxy defence in a lot of fraud stacks. Volume came back this week, but the mix did not flip: newly seen proxy exits were still almost all off datacenter space. The datacenter slice ticked up from 2.2% in the last posted recap to 2.4%, still in the flat ~2% band of recent weeks.

Between 12–18 November 2026 (UTC), the IP99 risk feed captured 10,705,894 proxy exit addresses it had not seen in the previous 7 days (+7% vs the previous 7 days). Source page: ip99.com/en/proxy-exits (generated 19 November 2026 UTC). The public page rolls daily over the last 7 complete UTC days; this is today's window, continuing the series after the 9–15 November recap (the two windows overlap on 12–15 November).

Against that last posted week (9,914,335 addresses), volume is up about +8.0%. The page's official +7% compares 12–18 November with 5–11 November — a clean adjacent week, and the same window as the 12 November recap (10,026,346). Much of the step-up versus the last posted recap is compositional: heavier 16–18 November days, led by 16 November (1,691,075), replaced lighter 9–11 November days such as 11 November (1,217,179). These figures describe what entered the 7-day mirror, not worldwide proxy supply.

What those addresses sit on

Usage type Addresses Share
Residential broadband 8,425,688 78.7%
Mobile network 1,431,350 13.4%
Gateway / NAT egress 514,896 4.8%
Datacenter 253,933 2.4%
Education network 50,157 0.5%
Unclassified 29,870 0.3%

About 63% were rated high-risk at capture; the rest still carried a proxy-class signal at a lower risk level.

IP99 counts an address as "captured" when it enters the feed with a proxy-class signal (proxy, VPN, dial-up pool, cloud phone, or hijacked shared proxy) and was not in the live 7-day risk mirror. Addresses that go quiet and return can be counted again. The page describes what 86 upstream collection sources delivered — not the total number of proxies in the world. An address is counted once per window. No address lists are published.

Why 2.4% still is not "mostly datacenter"

If your defence is mainly an ASN or hosting blocklist, you are still looking past most of the problem. Residential broadband alone is 78.7% of this week's capture window. Mobile adds another 13.4%. Together that is 92% of newly seen proxy exits sitting in access networks where real customers also live.

The datacenter row moved from 2.2% in the last two posted recaps (5–11 Nov and 9–15 Nov) to 2.4% here, still close to late October's 2.0–2.2% band after the brief rise to 3.6% (29 Oct–4 Nov) and 3.2% (4–10 Nov). A hosting-only blocklist still misses most of the supply-side traffic that shows up as residential or mobile exits.

Usage type answers "what kind of network is this," not "is this address currently sold as an exit." Rotating residential proxies and dial-up pools are designed to sit in DYN / MOB space. Detecting them needs supply-side observation with an age attached, not a lasting network-type label. See rotating residential proxies and detect proxy / VPN.

Which networks show up most

Top ASNs in this week's capture window (by distinct address count):

ASN Addresses Share
AS4134 947,700 8.9%
AS4837 707,646 6.6%
AS9808 358,103 3.3%
AS6057 276,201 2.6%
AS55836 259,403 2.4%

AS4134, AS4837, and AS9808 remain the three largest contributors. The top-five change is fifth place: AS55836, eighth in the last two posted recaps (1.6% on 5–11 Nov, 1.8% on 9–15 Nov) and fifth in the 29 Oct–4 Nov recap (2.2%), is fifth again at 2.4%. AS36903, fifth last week (2.2%), sits seventh (2.0%). Large access networks appearing here are a clue about where proxy supply is being sold from — not a claim that every customer on those networks is a proxy. Ban a whole ASN and you take out legitimate home and mobile users. Prefer per-address evidence. ASN pages: ip99.com/en/asn/.

Day-to-day shape

New-to-mirror captures that day, and how many were high-risk at capture:

Day (UTC) Captured High-risk at capture
2026-11-12 1,576,485 1,000,400
2026-11-13 1,606,501 994,471
2026-11-14 1,422,246 896,773
2026-11-15 1,253,419 804,951
2026-11-16 1,691,075 1,032,943
2026-11-17 1,596,167 983,685
2026-11-18 1,560,001 1,003,907

Volume was up on the official comparison (+7% vs 5–11 November). Upstream sources sat at 86, up from 60 in the last posted recap — a large jump the page reports, not a cause this article will invent. Daily capture stayed in a 1.25–1.69M band, lightest on 15 Nov (1,253,419) and heaviest on 16 Nov (1,691,075). That is still not a shape that makes a static list look "caught up."

How to use this without over-blocking

  1. Branch on evidence, not network type. Residential is where your customers live; do not hard-block on DYN / MOB alone.
  2. Step up, don't stop. A fresh proxy-class signal on signup, a coupon claim, or a login from a new device justifies a second factor or a delay — not an immediate ban of a whole building's shared egress.
  3. Read the age. evidence_state: "stale" means the address was in supply before and is not now; treat it as a weak prior. Re-query at decision time rather than caching a verdict for days. Score 0 with evidence_state: "none" is not a clean bill of health — it means no current verifiable evidence is on file.
  4. Keep the timestamp. Store computed_at with the decision so a disputed block can be traced back to when the observation was made.

IP99 by ThreatHunter publishes the free weekly aggregate so you can see the mix without buying a feed. The score on a single lookup is the freshness and strength of verifiable risk evidence — not a probability that the address "is an attacker." Do not treat IP99 scores as identical to every other ThreatHunter dataset you may have seen elsewhere; this article only claims what the public proxy-exits page and free /v1/ip document.

Check one yourself

Free, no signup:

curl https://ip99.com/v1/ip/8.8.8.8
Enter fullscreen mode Exit fullscreen mode

Look at evidence_state (active / stale / none), risk.score, and the proxy-class signals. Verified 19 November 2026, 13:12 UTC: this public DNS address still shows a hosting network attribute with score 0, evidence_state: "none", and usage_type IDC — a network fact, not a proxy verdict. The same lookup is available to agents as an MCP tool at https://ip99.com/mcp (ip_risk_lookup). Field notes live on the API docs.

Try it on ip99.com, keep an eye on this week's proxy exits, and wire the free OpenAPI at ip99.com/openapi.json for Postman or codegen.

Top comments (0)