Skip to content
CVE-2026-17132

IBM API Connect SQL Injection (CVE-2026-17132)

Detects exploitation attempts and indicators of CVE-2026-17132, a SQL injection vulnerability (CWE-89) in IBM API Connect versions 12.1.0.0 through 12.1.1.2. A remote authenticated attacker can execute arbitrary SQL commands due to improper neutralization of special elements used in an SQL command. This detection surfaces anomalous SQL syntax in authenticated API Connect requests, backend database errors indicative of injection, and post-exploitation database activity. CVSS 9.9 (critical).

Vulnerability Intelligence

Theoretical

Affected Software

Vendor
ibm
Product
api_connect
Versions
>= 12.1.0.0, < 12.1.1.3

Weakness (CWE)

Timeline

Disclosed
November 7, 2026

References & Proof of Concept

CVSS

9.9
Critical (9.0–10)

What is CVE-2026-17132 IBM API Connect SQL Injection (CVE-2026-17132)?

IBM API Connect SQL Injection (CVE-2026-17132) (CVE-2026-17132) maps to the Initial Access and Collection and Credential Access tactics — the adversary is trying to get into your network in MITRE ATT&CK.

This page provides production-ready detection logic for IBM API Connect SQL Injection (CVE-2026-17132), covering the data sources and telemetry it touches: IIS/Web server logs, Reverse proxy logs, IBM API Connect gateway logs. The queries below are rated critical severity at medium confidence, and ship for 7 SIEM platforms — KQL, SPL, Elastic, QRadar, Sumo, YARA-L, LogScale.

MITRE ATT&CK

Tactic
Initial Access Collection Credential Access
Microsoft Sentinel / Defender
kusto
// df00tech DF00-fddc63b249 · https://df00tech.com/detections/CVE-2026-17132 · licence: https://df00tech.com/detections/licence
let sqliPatterns = dynamic(["' or '1'='1", "union select", "'--", "';", "waitfor delay", "xp_cmdshell", "information_schema", "sleep(", "benchmark(", "/*!", "concat(", "0x", "convert(int"]);
W3CIISLog
| where csUriStem has_any ("/api/", "/apiconnect/", "/manager/", "/consumer-api/", "/analytics/")
| extend decodedQuery = tolower(url_decode(coalesce(csUriQuery, "")))
| extend decodedStem = tolower(url_decode(csUriStem))
| where decodedQuery has_any (sqliPatterns) or decodedStem has_any (sqliPatterns)
| where scStatus in ("200", "500", "302")
| project TimeGenerated, cIP, csUserName, csMethod, csUriStem, csUriQuery, scStatus, sPort, Computer
| summarize count(), AttemptStart=min(TimeGenerated), AttemptEnd=max(TimeGenerated), Stems=make_set(csUriStem, 20) by cIP, csUserName, Computer
| where count_ >= 1

Identifies authenticated HTTP requests to IBM API Connect endpoints containing SQL injection meta-characters and keywords in the URL path or query string, captured in W3C IIS logs.

critical severity medium confidence

Data Sources

IIS/Web server logs Reverse proxy logs IBM API Connect gateway logs

Required Tables

W3CIISLog

False Positives

  • Legitimate API clients submitting data containing SQL-like strings (e.g. analytics queries, search terms with the word 'select').
  • Security scanners and authorized penetration tests generating injection payloads against the appliance.
  • Developers or QA testing API filter parameters that legitimately contain SQL keywords such as 'union' or 'concat'.

Query fingerprint DF00-fddc63b249 · licence

Sigma logsource & community rules

The detection logic for IBM API Connect SQL Injection (CVE-2026-17132) (CVE-2026-17132) above is provided in a vendor-neutral form so you can deploy it on any SIEM. The same logic is shipped here as native KQL (Microsoft Sentinel / Defender), SPL (Splunk), Elastic (Elastic Security (EQL)), QRadar (IBM QRadar (AQL)), Sumo (Sumo Logic CSE), YARA-L (Google Chronicle / SecOps), LogScale (CrowdStrike LogScale (CQL)) queries. In Sigma terms, this detection targets the following logsource:

logsource:
  category: network_connection
  product: windows

Browse the community-maintained Sigma rules for this technique:


Testing Methodology

Validate this detection against 3 adversary techniques from Atomic Red Team. Each test below lists the behaviour to exercise and the telemetry you should expect to see. Executable commands and cleanup steps are further down this page.

  1. Test 1 UNION-based SQL injection probe against API Connect consumer API

    Expected signal: IIS/gateway access log entry with decoded 'union select' and 'FROM users' in the URI for an authenticated user.

  2. Test 2 Time-based blind SQL injection (pg_sleep)

    Expected signal: Access log entry containing 'pg_sleep' / stacked-query syntax plus an elevated backend response latency.

  3. Test 3 information_schema enumeration attempt

    Expected signal: W3CIISLog entry with decoded 'information_schema.tables' and UNION SELECT on an /api/ endpoint.


Response Playbook

Triage

  1. Confirm the targeted host is running IBM API Connect in the affected range (12.1.0.0 through 12.1.1.2); check the deployed build via the Cloud Manager 'About' page or `apicup version`.
  2. Identify the authenticated principal (csUserName/user field) behind the flagged requests, SQL injection here requires authentication, so determine whether the account is a legitimate consumer, a service account, or a compromised credential.
  3. Decode and inspect the full request URI and body to distinguish a genuine injection payload (e.g. UNION SELECT, information_schema enumeration, stacked queries) from benign data that merely contains SQL keywords.
  4. Correlate the request timestamps against backend database (DataPower/Postgres/management DB) error logs and slow-query logs for the same window to confirm the payload reached the database layer.

Containment

  1. Revoke or disable the implicated API Connect user account/API key and invalidate active sessions and OAuth tokens for that principal.
  2. Apply IBM's fix by upgrading API Connect to 12.1.1.3 or later; until then, restrict management/consumer API access to trusted networks via gateway ACLs or WAF rules that block SQL meta-characters.
  3. If exploitation is confirmed, isolate the affected management database and rotate its credentials, then review and restrict the DB account privileges used by API Connect.

Evidence Collection

  1. Preserve IIS/gateway access logs, API Connect management and analytics logs, and database query/error logs covering the attack window.
  2. Capture a forensic snapshot of the management database and API Connect configuration to identify any attacker-created or modified records.
  3. Export the full decoded request payloads, source IPs, user agents, and the authenticated identity associated with the flagged activity.

Escalation Criteria

  • ! Escalate to incident response if database error logs, UNION-based data in responses, or exfiltrated rows confirm that injected SQL executed successfully.
  • ! Escalate if the implicated account shows signs of compromise (impossible-travel logins, credential stuffing, or privilege changes) or if sensitive data (credentials, API keys, PII) was accessed.
  • ! Escalate to the platform owner if the affected API Connect version cannot be patched promptly and is internet-exposed.

Investigation Guide

Forensic Artifacts

  • > IIS/gateway W3C access logs showing SQL meta-characters in API Connect URIs
  • > API Connect management and analytics service logs
  • > Backend database (Postgres/management DB) query and error logs, slow-query logs
  • > Authentication/OAuth token issuance logs for the implicated principal

Tuning Guidance

Baseline normal API Connect consumer traffic first: identify endpoints and parameters that legitimately contain SQL-like tokens (search filters, analytics expressions) and exclude them or require multiple distinct injection markers before alerting. Reduce noise from authorized scanners by whitelisting their source IPs. Raise fidelity by correlating request-time with backend DB error responses or HTTP 500s, and by alerting only on authenticated principals since the vulnerability requires authentication. Prioritise alerts where the same user/IP shows repeated or escalating injection attempts.


Hunting Queries

Hunts for database enumeration and time-based blind SQL injection patterns against API Connect endpoints grouped by source and authenticated user.

Hunting — KQL
kql
// df00tech DF00-fddc63b249 · https://df00tech.com/detections/CVE-2026-17132 · licence: https://df00tech.com/detections/licence
W3CIISLog | where csUriStem has_any ("/api/","/apiconnect/","/manager/","/consumer-api/") | extend q=tolower(url_decode(coalesce(csUriQuery,""))) | where q has_any ("information_schema","union select","pg_sleep","waitfor delay") | summarize count() by cIP, csUserName, bin(TimeGenerated, 1h)
Hunting — SPL
spl
``` df00tech DF00-fddc63b249 · https://df00tech.com/detections/CVE-2026-17132 · licence: https://df00tech.com/detections/licence ```
index=web (sourcetype=ms:iis OR sourcetype=apiconnect:access) uri_path="*/api*" | eval p=lower(urldecode(uri_query)) | where match(p,"(?i)(information_schema|union\s+select|pg_sleep|waitfor\s+delay)") | stats count by src_ip, user

Atomic Red Team Tests

Test 1 UNION-based SQL injection probe against API Connect consumer API
linux

Sends an authenticated request with a UNION SELECT payload in a query parameter to simulate CVE-2026-17132 exploitation (lab only).

Command

bash
curl -s -k -H "Authorization: Bearer $APIC_TOKEN" "https://apic-lab.local/consumer-api/products?filter=1'%20UNION%20SELECT%20username,password%20FROM%20users--"

Cleanup

bash
echo 'No persistent artifact created by probe; clear any test records if the query succeeded.'

Expected Telemetry

IIS/gateway access log entry with decoded 'union select' and 'FROM users' in the URI for an authenticated user.

Expected Detection

KQL/SPL/EQL rules fire on the UNION SELECT pattern against a /consumer-api/ endpoint.

Test 2 Time-based blind SQL injection (pg_sleep)
linux

Issues a time-based blind SQLi payload using a database sleep function to simulate inference-based exploitation (lab only).

Command

bash
curl -s -k -H "Authorization: Bearer $APIC_TOKEN" "https://apic-lab.local/manager/apis?name=x';SELECT%20pg_sleep(10)--"

Cleanup

bash
echo 'No cleanup required; time-based payload leaves no data change.'

Expected Telemetry

Access log entry containing 'pg_sleep' / stacked-query syntax plus an elevated backend response latency.

Expected Detection

Detection matches the sleep()/waitfor delay time-based injection signatures on an API Connect endpoint.

Test 3 information_schema enumeration attempt
windows

Attempts to enumerate database structure via information_schema through an authenticated API Connect parameter (lab only).

Command

powershell
powershell -c "$h=@{Authorization='Bearer '+$env:APIC_TOKEN}; Invoke-WebRequest -UseBasicParsing -Uri \"https://apic-lab.local/api/catalogs?q=1' UNION SELECT table_name,1 FROM information_schema.tables--\" -Headers $h"

Cleanup

powershell
Remove-Item Env:\APIC_TOKEN -ErrorAction SilentlyContinue

Expected Telemetry

W3CIISLog entry with decoded 'information_schema.tables' and UNION SELECT on an /api/ endpoint.

Expected Detection

Detection fires on the information_schema enumeration pattern for the authenticated principal.

Related Detections

Browse by tactic

Full ATT&CK matrix Search all detections