A flaw in the libsoup library allows an attacker who has influence over the value used to set the Content‑Type header to inject a CRLF sequence due to lack of proper input sanitization in the `soup_message_headers_set_content_type()` function. The inserted CRLF characters create additional, arbitrary header‑value pairs in the HTTP response, which can lead to HTTP header injection and response‑splitting attacks. The vulnerability is identified as CWE‑93 and may be exploited by manipulating the outgoing HTTP headers that the application generates.

The issue affects Red Hat Enterprise Linux systems from version 6 through 10 that include an unpatched copies of libsoup. No specific sub‑versions are cited, so any Red Hat installation that links against a libsoup version before the fix is potentially vulnerable.

The CVSS score of 3.9 indicates a low‑to‑moderate severity, and the EPSS exploitation probability is reported as less than 1 %. The vulnerability is not listed in the CISA Known Exploited Vulnerabilities catalog. The likely attack requires an application that forwards user‑controlled or otherwise untrusted content‑type values to the `soup_message_headers_set_content_type()` call, enabling the attacker to inject CRLF characters and alter the structure of the HTTP response headers.