CISA added three vulnerabilities to its Known Exploited Vulnerabilities catalog on 4 August 2026, and the headline entry is a critical flaw in IBM Langflow, the low-code tool teams use to wire up AI agents.
The three additions are CVE-2026-9198 in IBM Langflow, CVE-2026-18556 in N-able N-central and CVE-2026-34486 in Apache Tomcat. CISA lists all three on evidence of active exploitation. Under Binding Operational Directive 26-04, federal civilian agencies are required to prioritize rapid remediation of high-risk KEV entries on publicly exposed assets, and to check whether attackers got in before the patch was applied.
Langflow is the most serious of the three
CVE-2026-9198 carries a CVSS base score of 9.8, the highest in the batch. The published NVD record describes an unauthenticated chain in Langflow OSS 1.0.0 through 1.10.0: a call to the /api/v1/auto_login endpoint mints a superuser token for any network caller that can reach it, and that token then unlocks /api/v1/validate/code, which executes supplied code. On a default Langflow deployment, that adds up to full remote code execution with no credentials needed. Because the record describes the token as being issued to any network caller, an instance reachable from the internet is exposed regardless of what credentials are configured on it.
Exposed Langflow servers have been a target for a while. In June, IntelFusions reported on attackers hijacking internet-facing Langflow instances to mine cryptocurrency, and in July researchers documented an AI agent running a ransomware attack against a Langflow host largely on its own. The KEV listing confirms a pattern rather than starting one.
Tomcat and N-able
CVE-2026-34486 in Apache Tomcat is scored 7.5. Its published record describes a missing-encryption issue introduced by the fix for CVE-2026-29146, which allows the EncryptInterceptor to be bypassed. Apache lists versions 11.0.20, 10.1.53 and 9.0.116 as affected, and recommends upgrading to 11.0.21, 10.1.54 or 9.0.117.
The N-able N-central authentication bypass will be familiar: IntelFusions covered it earlier this week when CISA warned that hackers were exploiting the remote monitoring platform's login bypass. CISA's alert gives only the CVE identifier, the product and the vulnerability class, and no severity score is published for it in the sources available here. The vendor advisory carries the detail.
What you should do
Patch all three, starting with anything internet-facing, and treat Langflow as the priority given the score and the lack of any authentication requirement. Federal civilian agencies fall under BOD 26-04 timelines. Everyone else should treat a KEV listing the same way, because it means exploitation has already been observed rather than merely predicted.
The additions were published in CISA's alert. Where severity scores appear above they come from the published CVE records; the KEV catalog itself does not publish scores.
This briefing is provided by IntelFusions for informational and defensive purposes only. It is based on sources assessed to be reliable at the time of writing, and analytic judgments carry the confidence levels indicated. Indicators of compromise are defanged; re-arm them only in controlled environments. IntelFusions is not affiliated with the organizations named and makes no warranty as to completeness or accuracy.