On this page (7 sections)
Two critical server-side request forgery (SSRF) vulnerabilities, CVE-2026-17298 and CVE-2026-17296, were published for IBM API Connect on 2026-11-07. The NVD record for CVE-2026-17298 says a remote authenticated attacker could abuse the flaw to obtain sensitive information.
If you run IBM API Connect in the affected version ranges, whether self-hosted or administered for a client, review IBM's support pages and plan an upgrade. The public record is short, so this article keeps what is confirmed apart from what is not.
What happened
The two CVE records share the same wording about the impact and the same affected ranges. Both were published on 2026-11-07 and both reference two IBM support pages: node 7295635 and node 7295637.
The scores differ. CVE-2026-17298 is listed at CVSS 9.9 and CVE-2026-17296 at 9.6, and both are rated critical. The research does not explain why two identifiers exist for the same description, so treat them as two distinct issues that affect the same versions until IBM's pages say otherwise.
What SSRF means in plain terms
Server-side request forgery lets an attacker make an application send network requests for them. Because the request comes from the server itself, it can reach places the attacker cannot reach directly, such as internal services, admin interfaces or cloud metadata endpoints. The typical outcome is disclosure of information that was never meant to leave the internal network, which fits the impact in these records: obtaining sensitive information.
API management platforms are sensitive places for this kind of bug. They usually sit between the internet and many backend services, and often hold credentials, tokens and configuration for those backends. The research does not say which API Connect component or feature is involved, and this article does not speculate about it.
Who is affected
| Release line | Affected versions |
|---|---|
| 10.0.8 | 10.0.8.0 through 10.0.8.12 |
| 12.1 | 12.1.0.0 through 12.1.1.2 |
The attacker must be authenticated, so practical exposure depends on who can log in to your API Connect environment. Count every account type: internal administrators, API developers, partner or third-party users, and any self-service sign-up. The larger and less trusted that group, the more urgent the upgrade. Authentication lowers the risk compared with an anonymous flaw but does not remove it, especially where credentials are shared, reused or phished.
How to check your exposure
Identify every API Connect deployment you operate, including test and staging, and note the exact version of each from your deployment records or the product's administration interface. Compare it with the ranges above, then read IBM's two support pages for fixed versions and interim mitigations, since the CVE records do not list them.
- Inventory all API Connect instances and record exact version numbers.
- Match each version against the 10.0.8.0 to 10.0.8.12 and 12.1.0.0 to 12.1.1.2 ranges.
- List all accounts that can authenticate, and note which are external or low-trust.
- Identify which internal services and credentials the API Connect hosts can reach on the network.
What to do now
- Open IBM's support pages (node 7295635 and node 7295637) and note the fixed versions and any workarounds.
- Schedule the upgrade or fix for every affected instance, starting with internet-facing ones.
- Remove unused or stale accounts, and enforce strong, unique credentials and multi-factor authentication where your setup supports it.
- Restrict outbound network access from API Connect hosts to the destinations they genuinely need, which limits what an SSRF flaw can reach.
- Make sure sensitive internal endpoints, such as cloud metadata services, are not reachable from the API Connect hosts unless required.
- Rotate secrets and tokens that the platform stores or can reach if you find signs of misuse.
- Review access and gateway logs for unusual requests from authenticated users to internal addresses.
Detection and monitoring
The research gives no indicators of compromise for these flaws, so detection has to rely on general SSRF signals. Look in your logs for authenticated users submitting URLs or hostnames that point to internal IP ranges, loopback addresses or metadata endpoints, and for unexpected outbound connections from the API Connect hosts to internal systems.
A website firewall can add filtering in front of web-facing applications, as described in what is a WAF, but it does not replace installing the vendor fix. For the public sites and certificates around your API endpoints, certificate and uptime monitoring can alert you to unexpected changes.
Timeline
| Date | Event |
|---|---|
| 2026-11-07 | CVE-2026-17298 and CVE-2026-17296 published in NVD, referencing IBM support pages 7295635 and 7295637 |
Sources
This article is based on the public sources below. Details can change as vendors and researchers publish more; check the original advisories before acting on version numbers.
- CVE-2026-17298: IBM API Connect 10.0.8.0 through 10.0.8.12 and 12.1.0.0 through 12.1.1.2 could allow a remote authenticated attacker to obtain sensitive information due to…
- CVE-2026-17296: IBM API Connect 10.0.8.0 through 10.0.8.12 and 12.1.0.0 through 12.1.1.2 could allow a remote authenticated attacker to obtain sensitive information due to…
