asn.dev

Autonomous system

AS12696 AXA-TECH Paris

Unclassified ARIN

IPv4 addresses 26.88K 20 prefixes
IPv6 prefixes 0 counted as prefixes, not addresses
Countries 1 mostly FR
Exchanges 1 recorded by PCH
Upstreams 12 0 downstream
Flagged prefixes 0 none

Summary

AXA Technology Services

France · AI-generated

finance enterprise business

An insurer's IT arm, whose netnames number their pools to match the addresses they hold.

What it's used for

  • Corporate networking for an insurance group. 26,368 addresses in nineteen IPv4 prefixes, 43 observed BGP neighbours, and no IPv6 announced.

Notable

  • Fifteen of the nineteen prefixes come out of 171.18.0.0, a single legacy block cut into pieces.
  • The netnames number their pools: AXA-GO-SEDC-POOL32 holds 171.18.32.0/21, AXA-GO-SEDC-POOL-80 holds 171.18.80.0/22, and AXA-GO-SEDC-SSN-PR-42 holds 171.18.42.0/23.
  • Others are AXA-TECH-NET1, AXA-TECH-SEDC-MR and a bare AXA.
  • The contact on every record is AXA GO abuse handling.
  • Eighteen sampled addresses were queried and none returned a name.

Interesting

  • The number at the end of each netname is the third octet of the block it describes, which makes the registry self-indexing: read the name, know the addresses. It works because one team allocated all of it out of one /16, and it would break the moment a block arrived from anywhere else.
  • 171.18 is legacy space from the era when a large company could ask for 65,536 addresses and be given them. The pieces announced here add up to well under half of that, so the rest is either used internally or sitting idle behind the aggregate.
  • 43 neighbours is a lot for an enterprise. An insurer connects to brokers, banks, clearing systems and its own offices, and each of those relationships is cheaper and more predictable as a direct session than as traffic crossing the public Internet.

This description is AI-generated, best-effort context — not authoritative, and provided without any guarantee of accuracy. Source, and how to suggest an edit.

Interconnection

Peering at a glance

Full topology · Graph

Internet exchanges

Recorded at 1 exchange in 1 country, largest first.

ExchangeLocationTheir address
France-IX ParisParis, France37.49.237.48

Transit providers

12 upstream, 0 downstream and 12 adjacent networks observed; most-used upstreams first.

ASNetworkPaths seen
AS32787Akamai Technologies, Inc.5,232
AS6453TATA COMMUNICATIONS (AMERICA) INC776
AS6939Hurricane Electric LLC553
AS20766Association "Gitoyen"79
AS24482SG.GS79
AS34927iFog GmbH79
AS37721Virtual Technologies & Solutions79
AS50628Leonix Telecom SAS79

Exchange presence is recorded rather than inferred: the directory collected by Packet Clearing House (https://www.pch.net/ixp/data) on 2026-10-17, licensed CC BY-NC-SA 3.0. Transit is inferred from observed BGP paths, never declared by either party.

Address space

Summary

Shape and geography

IPv4 20 prefixes · 26.88K addresses IPv6 0 prefixes

CountryNamePrefixes
FRFrance20

Registry

Allocation record

ARIN · via Atlas (registry + RDAP)
Registry
ARIN
Handle
AS12696
Registry name
AXA-TECH
Registrant
AS12696-MNT
Abuse contact
abuse@colt.net
Registered
1970-01-01
Last changed
2025-01-27
Allocation
allocated

Routing security

RPKI and IRR

Full detail
RPKI
0% valid · 0 valid, 118 with no ROA
ROAs
0 authorising AS12696
ASPA
no ASPA record
IRR
100% registered · 118 of 118 announcements
Route objects
146 naming AS12696

Reputation

Classification and threat

Unclassified Not in the curated lists. Most of the ~80,000 allocated ASNs are not.

Spamhaus DROP

Not on ASN-DROP The network as a whole is not listed.

No DROP prefixes None of its announced space is on the DROP prefix list.

Spamhaus DROP ("Don't Route Or Peer") lists space and networks run by, or hijacked for, spam and cybercrime operations: 442 networks and 1,783 prefixes, as published 2026-10-17. A short, high-confidence list, so a match is worth taking seriously, and absence from it says little on its own.

FBI IC3

No IC3 indicators Nothing from the FBI IC3 feed.

IC3 threat data is alpha. Absence of an indicator is not evidence of good behaviour, only of absence from this one feed.

Elsewhere

Cross-check