← Back to Feed CACHED · 2026-07-31 06:03:22 · CACHE_KEY CVE-2026-48449
CVE-2026-48449 · CWE-863 · Disclosed 2026-07-30

Adobe Campaign Classic

ASSESSED — NOISGATE V0.5
Vendor
Reassessed
Verdict:
Do you agree?
01 · The Real Story

Someone left the keys in the ignition of your marketing broadcast truck, and it's parked on the public street

CVE-2026-48449 is an Incorrect Authorization (CWE-863) vulnerability in Adobe Campaign Classic (ACC) v7, affecting on-premise deployments running version 7.4.3 build 9396 and earlier. The flaw allows an unauthenticated remote attacker to bypass authorization checks entirely and achieve arbitrary code execution in the context of the current user — with scope change, meaning the attacker can pivot beyond the vulnerable component. ACC's architecture includes an Apache Tomcat-based web module and tracking/redirection server that are internet-facing by design (placed in the DMZ to handle email click-tracking, landing pages, and unsubscribe links). The affected surface is therefore directly reachable from the internet in most on-premise deployments.

Adobe rated this a perfect 10.0 — the maximum possible CVSS score — and that rating is *technically accurate* for the attack mechanics: network-reachable, no auth, no user interaction, full CIA impact with scope change. However, noisgate downgrades to HIGH 9.0 because ACC is a niche line-of-business marketing platform, not a high-value infrastructure component (it is not an identity provider, domain controller, hypervisor, or backup server). The on-premise installed base is a fraction of all enterprises, Adobe-hosted and hybrid-hosted instances are already patched server-side, there is no public PoC, no active exploitation, and EPSS sits at a low 0.54%. That said, the downgrade is modest — ACC servers hold massive customer PII (email addresses, demographics, purchase history), so a breach carries significant regulatory and reputational consequences. With related Adobe ColdFusion CVEs (CVE-2026-48282) already under active exploitation, threat-actor interest in Adobe product vulns is elevated right now.

"Unauth RCE in internet-facing marketing infra — HIGH for niche but potent blast radius"
02 · The Attack Path

4 steps from start to impact.

STEP 01

Identify exposed ACC web module

The attacker scans for internet-facing ACC instances. ACC's tracking/redirection server (nlserver webmdl) and Apache Tomcat entry point (typically port 8080 or reverse-proxied on 443) are placed in the DMZ by design to serve email tracking pixels, landing pages, and unsubscribe workflows. Shodan/Censys/FOFA fingerprinting can identify ACC instances via HTTP response headers and default paths.
Conditions required:
  • ACC on-premise deployment exists
  • Tracking/web module is internet-facing (standard architecture)
Where this breaks in practice:
  • ACC is a niche product — limited installed base compared to Exchange, SharePoint, or ColdFusion
  • Some orgs may restrict ACC behind VPN or WAF with strict allowlisting
Detection/coverage: Shodan/Censys can identify exposed ACC instances. WAF logs showing unusual requests to ACC SOAP/API endpoints should trigger alerts.
STEP 02

Craft unauthenticated authorization bypass request

The attacker sends a crafted HTTP request to the ACC web endpoint that exploits the incorrect authorization logic. Because PR:N (no privileges required) and UI:N (no user interaction), the request bypasses all authentication and authorization gates. The exact payload is not publicly documented yet — no PoC exists — but the CWE-863 class typically involves missing or flawed access-control checks on privileged API endpoints or SOAP methods.
Conditions required:
  • Network access to the ACC web module
  • Knowledge of the vulnerable endpoint (not yet public)
Where this breaks in practice:
  • No public PoC or exploit code as of 2026-07-31
  • Reverse-engineering the patch diff to identify the vulnerable code path requires effort
  • WAF rules tuned for ACC SOAP traffic may block anomalous requests
Detection/coverage: SOAP/API request anomaly detection. IDS signatures will likely follow once the patch diff is analyzed.
STEP 03

Achieve arbitrary code execution with scope change

Successful exploitation grants the attacker code execution in the context of the ACC process user. The CVSS S:C (scope change) indicates the attacker can affect resources beyond the vulnerable component's security scope — this likely means escaping from the ACC application context to the underlying OS or pivoting to connected databases and internal systems. ACC runs on Windows or Linux servers and typically connects to Oracle, PostgreSQL, or MSSQL databases containing campaign data and customer PII.
Conditions required:
  • Successful authorization bypass from Step 2
Where this breaks in practice:
  • Application-level sandboxing or containerization could limit OS-level access
  • Database credentials may be stored with least-privilege, limiting lateral scope
Detection/coverage: EDR on the ACC server should detect anomalous process spawning from the Tomcat/nlserver process tree. Database audit logs would show unusual queries.
STEP 04

Exfiltrate customer PII and pivot internally

With OS-level access on the ACC server, the attacker extracts the marketing database containing customer PII (email addresses, names, demographics, behavioral data, purchase history). The ACC server's network position in the DMZ may also provide a pivot point into internal networks if segmentation is weak. The attacker can also abuse ACC's email delivery infrastructure to send phishing campaigns from a trusted sender reputation.
Conditions required:
  • Code execution on ACC server
  • Database credentials accessible from ACC config files
  • Weak network segmentation between DMZ and internal network
Where this breaks in practice:
  • Proper DMZ segmentation limits lateral movement
  • DLP controls may detect bulk PII exfiltration
  • Email delivery abuse would be caught by outbound email monitoring
Detection/coverage: DLP, NetFlow anomaly detection, outbound email volume alerts, SIEM correlation of ACC server activity with unusual internal connections.
03 · Intelligence Metadata

The supporting signals.

In-the-wild exploitationNot observed. No reports of active exploitation of CVE-2026-48449 as of 2026-07-31. However, related Adobe CVE-2026-48282 (ColdFusion) is being actively exploited per Field Effect, indicating heightened threat-actor interest in this patch cycle.
Proof of conceptNone public. No PoC repositories (SecureWithUmer/CVE-2026-PoCs, 0xMarcio/cve, pwnfuzz/cve-pocs) contain exploit code for this CVE as of 2026-07-31.
EPSS score0.00542 (~0.54%) — low probability of exploitation in the next 30 days per FIRST EPSS model. Expect this to rise if a PoC drops.
KEV statusNot listed in CISA Known Exploited Vulnerabilities catalog. Related CVE-2026-48282 (ColdFusion) has been added to KEV.
CVSS vectorCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H10.0. Network-reachable, no auth, no interaction, scope change, full CIA. Maximum possible severity rating.
Affected versionsAdobe Campaign Classic (ACC) v7: 7.4.3 build 9396 and earlier (on-premise and on-premise components of hybrid deployments). Adobe-hosted instances are already patched.
Fixed versionACC v7: 7.4.3 build 9397 per Adobe Security Bulletin APSB26-69.
Scanning/exposure dataACC tracking servers are architecturally placed in the DMZ and are internet-facing by design (email tracking, landing pages, unsubscribe links). No specific Shodan/Censys enumeration data available yet for this CVE.
Disclosure date2026-07-30 (Adobe advisory). Adobe Priority Rating: 1 (potential for imminent exploitation).
Reporting researcherNot publicly attributed. Disclosed via Adobe's regular security bulletin process.
04 · The Call

noisgate verdict.

Final Verdict
DOWNGRADED to HIGH (9.0/10)

Downgraded from CRITICAL 10.0 to HIGH 9.0 because ACC is a niche line-of-business marketing server — not a high-value infrastructure role (not an identity provider, domain controller, hypervisor, or backup system) — which limits the affected population and caps the blast radius at customer-PII exfiltration rather than fleet-scale compromise. The single most decisive factor preventing a full CRITICAL rating is the limited installed base: only enterprises running ACC v7 on-premise are affected, and many ACC customers use Adobe-hosted or hybrid deployments that are already patched server-side.

HIGH Vulnerability severity and technical impact assessment
MEDIUM Exact attack surface and exploitability (no PoC available to validate)
LOW Real-world exposure count of on-premise ACC instances

Why this verdict

  • Unauthenticated network RCE with scope change: The CVSS vector is maximum severity — no auth, no interaction, scope change to OS/database. Any attacker with network access to the ACC web module gets code execution. This alone justifies HIGH or above.
  • Internet-facing by design: ACC's tracking/redirection server and web module are architecturally placed in the DMZ. Unlike many enterprise apps that *can* be segmented away, ACC *must* be reachable from the internet for email tracking and landing pages, making the attack surface directly exposed.
  • Role multiplier: ACC occupies a *line-of-business marketing server* role. In this role, the blast radius is host → tenant → customer PII database (mass data egress of regulated personal data). It does NOT reach domain → fleet → supply-chain scale. ACC is not canonically a high-value-role component (not IdP, DC, hypervisor, backup, CI/CD, or network edge). The floor rule does not mandate CRITICAL.
  • Temporal friction: No public PoC, no active exploitation, EPSS at 0.54%. These are temporal factors that could shift rapidly — especially given that related Adobe ColdFusion CVEs from the same patch cycle ARE being exploited, indicating active attacker interest in Adobe product vulnerabilities right now.
  • Reduced population: Only on-premise ACC v7 deployments are affected. Adobe-hosted and hybrid-hosted instances have been patched server-side. This narrows the vulnerable population significantly.

Why not higher?

A CRITICAL rating would be warranted if ACC were a high-value infrastructure component where compromise leads to fleet-scale or identity-scale impact (domain controller, hypervisor, IdP, backup). ACC is a marketing automation server — significant for PII exposure and regulatory risk, but the blast radius is bounded to the ACC server, its connected databases, and potentially the DMZ network segment. The absence of active exploitation, public PoC, and KEV listing further supports keeping this one notch below CRITICAL.

Why not lower?

Dropping below HIGH would be irresponsible given the perfect CVSS 10.0 mechanics: unauthenticated, network-reachable, no user interaction, with scope change. ACC servers are internet-facing by design and contain regulated customer PII. Related Adobe CVEs are already under active exploitation, and Adobe assigned Priority Rating 1. The temporal calm (no PoC, no exploitation) is fragile and could break any day.

05 · Compensating Control

What to do — in priority order.

  1. Restrict network access to ACC web endpoints immediately — Place ACC behind a WAF with strict allowlisting of known IP ranges (ESPs, internal admin IPs). Block all unauthenticated access to SOAP/API endpoints. If tracking functionality must remain public, restrict to only the /r/ redirection and /res/ resource paths. Deploy within the noisgate HIGH mitigation SLA of 30 days — but given the 10.0 CVSS and Adobe Priority 1 rating, aim for days not weeks.
  2. Apply ACC v7.4.3 build 9397 patch — This is the vendor fix. On-premise ACC administrators should apply build 9397 as the definitive remediation. Test in staging first given ACC's tight coupling with marketing databases and delivery workflows.
  3. Enable EDR monitoring on ACC servers — Ensure EDR agents are deployed and actively monitoring the ACC application server. Alert on anomalous child processes spawned by Tomcat/nlserver, unexpected outbound connections, and bulk database queries.
  4. Audit ACC database credential scope — Verify that ACC's database connection uses least-privilege credentials. The scope-change aspect of this CVE may allow OS-level pivot to database access — limiting DB permissions reduces blast radius.
  5. Monitor outbound email volume from ACC — A compromised ACC server could be weaponized to send phishing emails from your trusted sender reputation. Set up alerts for abnormal email volume or recipient patterns.
What doesn't work
  • MFA on ACC console login — This vulnerability is *unauthenticated* (PR:N). MFA protects the admin console but has zero effect on the authorization bypass attack path, which hits the web module directly.
  • Patching Adobe-hosted instances — If you're on Adobe-hosted or hybrid-hosted ACC, the server-side components are already patched by Adobe. This CVE only affects on-premise deployments.
  • Generic IDS/IPS signatures — No public PoC or exploit code exists yet, so signature-based detection has no patterns to match. Behavioral/anomaly detection is the right approach until specific signatures are released.
06 · Verification

Crowdsourced verification payload.

Run this script on each on-premise ACC server (Linux) or adapt the nlserver path for Windows. Requires shell access to the ACC host with read permissions on the nlserver binary. Example: chmod +x check_acc_cve.sh && ./check_acc_cve.sh

noisgate-verify.sh
BASHREAD-ONLYSAFE
#!/bin/bash
# CVE-2026-48449 / APSB26-69 — Adobe Campaign Classic build check
# Run on the ACC application server. No elevated privileges needed.
# Outputs: VULNERABLE / PATCHED / UNKNOWN

PATCHED_BUILD=9397

# Try to get ACC build number from nlserver
if command -v nlserver &>/dev/null; then
  VERSION_OUTPUT=$(nlserver -version 2>&1)
elif [ -x /usr/local/neolane/nl*/bin/nlserver ]; then
  VERSION_OUTPUT=$(/usr/local/neolane/nl*/bin/nlserver -version 2>&1)
else
  echo "UNKNOWN — nlserver binary not found. Verify this is an ACC server."
  exit 2
fi

# Extract build number (format: "Adobe Campaign Classic vX.Y.Z build NNNNN")
BUILD_NUM=$(echo "$VERSION_OUTPUT" | grep -oP 'build\s+\K[0-9]+')

if [ -z "$BUILD_NUM" ]; then
  echo "UNKNOWN — could not parse build number from: $VERSION_OUTPUT"
  exit 2
fi

echo "Detected ACC build: $BUILD_NUM (patched build: $PATCHED_BUILD)"

if [ "$BUILD_NUM" -ge "$PATCHED_BUILD" ]; then
  echo "PATCHED — ACC build $BUILD_NUM >= $PATCHED_BUILD. CVE-2026-48449 is remediated."
  exit 0
else
  echo "VULNERABLE — ACC build $BUILD_NUM < $PATCHED_BUILD. Apply ACC v7.4.3 build 9397 immediately."
  exit 1
fi
07 · Bottom Line

If you remember one thing.

TL;DR
If you run Adobe Campaign Classic v7 on-premise, this is your priority for the week. CVE-2026-48449 is unauthenticated RCE on an internet-facing server — the vendor gave it a 10.0 for good reason. noisgate rates it HIGH (9.0) due to the niche installed base, but for your org specifically, treat it as an emergency. Under the noisgate mitigation SLA for HIGH, deploy compensating controls (WAF restrictions on ACC web endpoints, EDR on ACC servers) within 30 days — but frankly, do it this week given the CVSS 10.0 and the fact that related Adobe ColdFusion CVEs are already being exploited in the wild. Apply the vendor patch (ACC v7.4.3 build 9397) under the noisgate remediation SLA of 180 days, but push for your next maintenance window. If you're on Adobe-hosted or hybrid-hosted ACC, your server-side components are already patched — verify with your Adobe TAM and move on. Monday morning: inventory your on-premise ACC instances, confirm which are internet-facing, apply WAF restrictions, and schedule the patch.

Sources

  1. Adobe Security Bulletin APSB26-69
  2. BleepingComputer — Adobe patches seven max severity ColdFusion, Campaign flaws
  3. SecurityWeek — Adobe Patches Critical ColdFusion, Campaign Classic Vulnerabilities
  4. The Hacker News — Adobe Patches 7 CVSS 10.0 Flaws
  5. Field Effect — Exploitation reported following Adobe security updates
  6. BeyondMachines — Adobe Patches Seven Maximum-Severity Flaws
  7. Canadian Centre for Cyber Security — Adobe advisory AV26-647
  8. Adobe Campaign Classic Architecture Documentation
Peer Review

What defenders are saying.

Submit a review attribution: handle + country only
0 flags selected · stored anonymously
Validation Results

Crowdsourced verification outputs.

Results submitted by users who ran the verification payload against their environment.