The vulnerability CVE-2026-41705 affects Spring AI's MilvusVectorStore#doDelete(List) implementation, allowing expression language injection via unsanitized document IDs. This enables an attacker to inject malicious filter expressions, potentially compromising confidentiality with limited impact on integrity and availability. Affected versions include Spring AI 1.0.0 through the latest 1.0.x and 1.1.0 through the latest 1.1.x. The CVSS 3.1 base score is 8.6, reflecting network attack vector, low attack complexity, no privileges required, no user interaction, and high confidentiality impact. The vendor recommends upgrading to versions 1.0.7 or greater and 1.1.6 or greater to remediate the issue. However, no explicit patch or vendor advisory content is provided in the input data.

Successful exploitation of this vulnerability can lead to unauthorized information disclosure due to expression language injection in the filter expressions used by the affected method. The CVSS score indicates high confidentiality impact, with limited impact on integrity and availability. There are no known exploits in the wild at this time.

Upgrade affected Spring AI versions to 1.0.7 or greater for the 1.0.x branch and 1.1.6 or greater for the 1.1.x branch as recommended. Patch status is not explicitly confirmed in the vendor advisory content provided; therefore, verify the availability of official patches or updates from the vendor. No alternative mitigations are specified.