~/f4n6 $ grep -r "eBanking Phishing Delivered Through IPv4-Mapped IPv6 Address, (Fri, Jun 19th)" ./investigations/ --include="*.md"

eBanking Phishing Delivered Through IPv4-Mapped IPv6 Address, (Fri, Jun 19th)

Jeff Davies 19 Jun 2026 7 min read

1. Executive summary

A phishing campaign targeting customers of Belfius, a major Belgian bank, has been observed using an IPv4-mapped IPv6 address notation (::ffff:) inside a URL to evade regex-based URL/domain extraction in security tooling. The literal IPv6 host ::ffff:5511:74be resolves to the IPv4 address 85.17.116.190 and contains no DNS record, defeating simple IOC-matching on domain names. The landing page redirects to hxxps://3439-aanmelden[.]verificatie[.]qzz[.]io/mon-belfius, a credential-harvesting kit impersonating Belfius. The risk to EMEA financial services is twofold: (1) end-user credential theft against Belgian retail and SMB banking customers, and (2) defensive-control blind spots where security stacks parse only IPv4 or extract domains via naive regex and therefore miss the malicious URL entirely. No CVE, no malware payload, and no threat-actor attribution are present in the source material; this advisory is treated as unconfirmed-attribution.

2. Regulatory framing

Article Trigger (the fact in this item) Practical impact
DORA Art. 17 Phishing email targeting bank customers constitutes an ICT-related incident requiring a documented management process. Financial entities must ensure their ICT-related incident management process covers phishing-driven credential theft and the IPv4-mapped IPv6 URL bypass technique.
DORA Art. 18 The campaign is a cyber threat requiring classification against the entity's taxonomy. Classify this as a phishing/credential-theft threat; tag the IPv4-mapped IPv6 URL pattern as a high-fidelity indicator for triage.
DORA Art. 19 A successful compromise of a customer or staff account would constitute a major ICT-related incident and trigger reporting to the competent authority. Pre-stage reporting templates and ensure the IPv4-mapped IPv6 URL and the qzz.io redirect host are captured as evidence in the report.
DORA Art. 28 The phishing kit is hosted on a third-party provider (qzz.io), creating third-party risk exposure. Assess concentration risk on free/personal hosting providers used for phishing infrastructure and ensure contractual/SLA coverage for takedown and abuse handling.
DORA Art. 29 Hosting on qzz.io is a potential concentration-risk vector for phishing infrastructure. Inventory exposure to qzz.io and similar low-cost hosting providers; ensure exit/containment paths exist.
DORA Art. 30 Third-party hosting of phishing kit requires contractual provisions for abuse handling and takedown. Verify that contracts with hosting/CDN providers include abuse-response SLAs covering credential-phishing infrastructure.
NIS2 Art. 21(2)(d) Supply-chain attack vector: phishing infrastructure hosted on third-party platform qzz.io. Apply supply-chain security measures: monitor and block qzz.io and subdomains where not business-justified; require vendors/hosters to respond to abuse within defined windows.
NIS2 Art. 23 A successful credential-theft incident would trigger incident-reporting obligations. Ensure incident-reporting workflows capture the IPv4-mapped IPv6 URL bypass as a notable technique in the initial notification.
UK NIS 2018 OES/RDSP duties apply where the phishing affects essential services (e.g., banking/payments). UK OES/RDSP entities must include this campaign in threat-intel feeds and ensure detection coverage for the IPv4-mapped IPv6 URL pattern.

3. Technical analysis & attack chain

The campaign uses a single-stage URL that combines an IPv4-mapped IPv6 literal with a redirect to a credential-harvesting kit. The technique is the headline: it weaponises RFC 4291 IPv4-mapped IPv6 address syntax to defeat regex-based URL/domain extraction in security tooling.

Attack chain (confirmed steps)

  1. Phishing email delivery. Email contains a malicious link. The phishing lure itself is described as "classic" and not detailed in the source.
  2. Malicious URL uses IPv4-mapped IPv6 literal. Link is hxxp://[::ffff:5511:74be]/kWC5PHA1. The [...] notation tells the URL parser the content is a literal IPv6 address (per RFC 3986 / RFC 6874).
  3. IPv4-mapped IPv6 expansion. The leading :: expands to 0000:0000:0000:0000:0000:ffff:5511:74be. The ::ffff: prefix is the IPv4-mapped IPv6 prefix defined in RFC 4291. The trailing two 16-bit hex groups 5511 and 74be are the four IPv4 octets in hex.
  4. Hex-to-decimal conversion. 0x55=85, 0x11=17, 0x74=116, 0xBE=190. The literal IPv6 host therefore resolves to 85.17.116.190.
  5. No DNS resolution required. Because the URL uses a literal IP, no DNS lookup is performed; this also means there is no DNS record to pivot on or block via DNS-layer controls.
  6. Redirect to credential-harvesting kit. The IPv4 host redirects to hxxps://3439-aanmelden[.]verificatie[.]qzz[.]io/mon-belfius, a page impersonating Belfius (Belgian bank) login.
  7. Credential capture. The phishing kit harvests Belfius banking credentials from victims.

Technical specifics that matter to a defender

  • Initial access vector: Phishing email with embedded URL. No attachment, no macro, no exploit.
  • Exploited component / "vulnerability": Defensive-control blind spot. The technique bypasses security tooling that extracts domains/IPs from URLs using simple regex. Any control that:
  • Matches only IPv4 dotted-quad patterns,
  • Strips or fails to parse [...] IPv6 literals,
  • Only blocks by domain name (FQDN/SLD), will miss this URL.
  • Vulnerability mechanism: RFC 4291 IPv4-mapped IPv6 address format (::ffff:a.b.c.d) embedded inside an IPv6 literal URL host. The URL parser correctly resolves to IPv4 85.17.116.190, but the textual representation in logs and IOC feeds is the IPv6 literal.
  • Payload / malware capabilities: None observed in source. The "payload" is a credential-harvesting web page impersonating Belfius.
  • Persistence: None observed. Single-session credential theft.
  • Privilege escalation: None observed.
  • Command-and-control: None observed. The phishing kit is hosted on qzz.io (a third-party hosting platform).
  • Lateral movement: None observed.
  • Data access / exfiltration: Belfius banking credentials harvested client-side via the phishing kit; exfiltration path not described in source.
  • Observed impact: Targeted credential theft against Belfius customers.

Caveats

  • The phishing email body, sender, and lure details are not described in the source ("classic" only).
  • The redirect chain beyond the second URL is not described.
  • The hosting provider qzz.io is referenced but not characterised; treat as unconfirmed third-party hosting.
  • No threat-actor attribution is provided in the source material; treat any attribution as unconfirmed.

4. Mitigation & containment

P1 — within 24 hours

  • Block the IPv4-mapped IPv6 URL at the proxy / web gateway. Add a literal-string block for hxxp://[::ffff:5511:74be]/kWC5PHA1 and the expanded form hxxp://[::ffff:85.17.116.190]/kWC5PHA1 in URL filtering policies.
  • Block the IPv4 destination at the firewall / NDR. Deny egress to 85.17.116.190 (port 80/443) at the perimeter.
  • Block the redirect destination. Add 3439-aanmelden.verificatie.qzz.io and the parent qzz.io (where not business-justified) to the DNS sinkhole and web-proxy block list.
  • Search email gateway logs for the literal IPv6 URL pattern (hxxp://[::ffff:) and the redirect URL across the last 30 days; quarantine any matching messages and notify recipients.
  • Hunt proxy / DNS logs for any successful connections to 85.17.116.190 or *.qzz.io paths matching mon-belfius or verificatie.

P2 — within 72 hours

  • Takedown request. Submit abuse notifications to the qzz.io provider and the IP owner of 85.17.116.190 requesting removal of the phishing kit and URL.
  • Force credential resets for any user observed connecting to either URL.
  • Update SIEM correlation rules to flag any URL containing the IPv4-mapped IPv6 prefix ::ffff: followed by hex characters, regardless of destination.
  • Review and harden URL-parsing logic in any in-house security tooling (regex-based IOC extractors, SOAR enrichment, email parsing) to correctly handle IPv6 literals and IPv4-mapped IPv6 addresses per RFC 4291.

P3 — within 7 days

  • User awareness push. Targeted warning to staff and (where applicable) customers about the Belfius-themed lure and the IPv4-mapped IPv6 URL technique.
  • Detection engineering. Roll out the Sigma rule in Section 6 across all log sources.
  • Third-party risk review. Confirm contractual abuse-response SLAs with hosting/CDN providers cover credential-phishing infrastructure takedown.

5. Indicators of compromise

Type Value Confidence Source
url hxxp://[::ffff:5511:74be]/kWC5PHA1 High SANS ISC diary 33090
ipv4 85.17.116.190 High SANS ISC diary 33090 (derived from ::ffff:5511:74be)
url hxxps://3439-aanmelden.verificatie.qzz.io/mon-belfius High SANS ISC diary 33090
domain 3439-aanmelden.verificatie.qzz.io High SANS ISC diary 33090
domain qzz.io Medium (parent hosting platform) SANS ISC diary 33090
url       hxxp://[::ffff:5511:74be]/kWC5PHA1
ipv4      85.17.116.190
url       hxxps://3439-aanmelden.verificatie.qzz.io/mon-belfius
domain    3439-aanmelden.verificatie.qzz.io
domain    qzz.io

6. Detection

Sigma rule — IPv4-mapped IPv6 URL in proxy / web logs

title: Web Request to IPv4-Mapped IPv6 Literal URL (Belfius Phishing)
id: 9c1f3b2a-7e4d-4a1b-8f3a-1b2c3d4e5f60
status: experimental
description: |
  Detects HTTP(S) requests where the URL host is an IPv4-mapped IPv6 address
  literal (RFC 4291 ::ffff: prefix) or contains the known Belfius phishing
  redirect path on qzz.io. Used by AT-2026-06-19-125.
author: Adverse Trace
date: 2026-06-19
reference: https://isc.sans.edu/diary/rss/33090
logsource:
  category: proxy
detection:
  selection_ipv4_mapped_literal:
    cs-uri-host|startswith: '['
    cs-uri-host|contains: '::ffff:'
  selection_known_ioc:
    cs-uri-host:

      - '3439-aanmelden.verificatie.qzz.io'
      - 'qzz.io'
    cs-uri-stem|contains: 'mon-belfius'
  condition: selection_ipv4_mapped_literal or selection_known_ioc
fields:

  - cs-uri-host
  - cs-uri-stem
  - c-ip
  - c-user
falsepositives:

  - Legitimate IPv4-mapped IPv6 URLs from internal services (rare; review)
level: high

Sigma rule — DNS query to phishing redirect domain

title: DNS Query to qzz.io Belfius Phishing Redirect Domain
id: 4d2e8a91-1c5b-4f6e-9a2d-3e4f5a6b7c80
status: experimental
description: |
  Detects DNS queries resolving to the qzz.io subdomain used to host the
  Belfius credential-phishing kit described in AT-2026-06-19-125.
author: Adverse Trace
date: 2026-06-19
reference: https://isc.sans.edu/diary/rss/33090
logsource:
  category: dns
detection:
  selection_query:
    query|contains:

      - '3439-aanmelden.verificatie.qzz.io'
      - 'verificatie.qzz.io'
  condition: selection_query
fields:

  - query
  - answer
  - src_ip
falsepositives:

  - None expected
level: high

7. Sources

  • SANS Internet Storm Center, "eBanking Phishing Delivered Through IPv4-Mapped IPv6 Address" by Xavier Mertens (@xme), 2026-06-19. https://isc.sans.edu/diary/rss/33090
  • RFC 4291, "IP Version 6 Addressing Architecture" (referenced for IPv4-mapped IPv6 prefix definition). https://www.rfc-editor.org/info/rfc4291/

8. Adverse Trace position

This is a phishing campaign, not a vulnerability — there is no CVE, no CVSS score, and no CISA KEV state to report. Severity is driven by the credential-theft impact against Belfius customers and the defensive-control bypass technique, which is the campaign's primary innovation. The IPv4-mapped IPv6 URL pattern is the high-fidelity detection signal: any URL whose host begins with [ and contains ::ffff: should be treated as suspect until reviewed, regardless of the destination IP. EMEA financial services clients should (1) confirm proxy / email-gateway coverage for the IOCs in Section 5, (2) audit in-house URL-parsing logic for IPv6-literal handling, and (3) push the Sigma rules in Section 6 to detection engineering for rollout. Adverse Trace will continue to monitor for reuse of the ::ffff: URL pattern against other EU banking brands and will update this advisory if additional infrastructure or attribution emerges.


Read the original source →

Published via PulseTrace — Adverse Trace threat intelligence.

Post this to LinkedIn
Formatting is converted automatically — headings, bullets, a link back & hashtags. Paste straight in.
J
Jeff Davies