Raw vector
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:HCVSS and EPSS are reproduced from their sources (NVD, FIRST EPSS). Risk Priority is our own derived reading, not an NVD score.
Summary
CVE-2023-39913 is a high-severity Improper Input Validation (CWE-20) vulnerability in Apache Uimaj. Its CVSS base score is 8.8 (High).
Operationally, exploitation aligns with the MITRE ATT&CK technique Junk Data (T1001.001); ranked in the top 27% of CVEs by exploit likelihood; it is not currently listed in the CISA KEV catalog.
OWASP Top 10 for Web (2025)
EU & UK References
- 🇪🇺 ENISA EUVD: EUVD-2023-2891
Vulnerability Data
Deserialization of Untrusted Data, Improper Input Validation vulnerability in Apache UIMA Java SDK, Apache UIMA Java SDK, Apache UIMA Java SDK, Apache UIMA Java SDK.This issue affects Apache UIMA Java SDK: before 3.5.0. Users are recommended to upgrade to version…
more
3.5.0, which fixes the issue. There are several locations in the code where serialized Java objects are deserialized without verifying the data. This affects in particular: * the deserialization of a Java-serialized CAS, but also other binary CAS formats that include TSI information using the CasIOUtils class; * the CAS Editor Eclipse plugin which uses the the CasIOUtils class to load data; * the deserialization of a Java-serialized CAS of the Vinci Analysis Engine service which can receive using Java-serialized CAS objects over network connections; * the CasAnnotationViewerApplet and the CasTreeViewerApplet; * the checkpointing feature of the CPE module. Note that the UIMA framework by default does not start any remotely accessible services (i.e. Vinci) that would be vulnerable to this issue. A user or developer would need to make an active choice to start such a service. However, users or developers may use the CasIOUtils in their own applications and services to parse serialized CAS data. They are affected by this issue unless they ensure that the data passed to CasIOUtils is not a serialized Java object. When using Vinci or using CasIOUtils in own services/applications, the unrestricted deserialization of Java-serialized CAS files may allow arbitrary (remote) code execution. As a remedy, it is possible to set up a global or context-specific ObjectInputFilter (cf. https://openjdk.org/jeps/290 and https://openjdk.org/jeps/415 ) if running UIMA on a Java version that supports it. Note that Java 1.8 does not support the ObjectInputFilter, so there is no remedy when running on this out-of-support platform. An upgrade to a recent Java version is strongly recommended if you need to secure an UIMA version that is affected by this issue. To mitigate the issue on a Java 9+ platform, you can configure a filter pattern through the "jdk.serialFilter" system property using a semicolon as a separator: To allow deserializing Java-serialized binary CASes, add the classes: * org.apache.uima.cas.impl.CASCompleteSerializer * org.apache.uima.cas.impl.CASMgrSerializer * org.apache.uima.cas.impl.CASSerializer * java.lang.String To allow deserializing CPE Checkpoint data, add the following classes (and any custom classes your application uses to store its checkpoints): * org.apache.uima.collection.impl.cpm.CheckpointData * org.apache.uima.util.ProcessTrace * org.apache.uima.util.impl.ProcessTrace_impl * org.apache.uima.collection.base_cpm.SynchPoint Make sure to use "!*" as the final component to the filter pattern to disallow deserialization of any classes not listed in the pattern. Apache UIMA 3.5.0 uses tightly scoped ObjectInputFilters when reading Java-serialized data depending on the type of data being expected. Configuring a global filter is not necessary with this version.
- CWE(s)
Related Threats
MITRE ATT&CK Enterprise Techniques
CVEs Like This One
Affected Assets
Mitigating Controls
Likely Mitigating Controls AI
Per-CVE control mapping for this CVE has not run yet; the list below is derived from the weakness types (CWEs) cited in the NVD entry.
Security testing and evaluation at multiple SDLC stages directly detects missing or flawed input validation, with the required remediation process ensuring fixes are applied.
Directly implements checks on information inputs to reject invalid data before processing.
Penetration testing supplies malicious serialized objects, detecting unsafe deserialization and supporting corrective actions.
Security testing and developer training directly verify and enforce proper input validation, reducing exploitability of injection and malformed-data weaknesses.
Untrusted serialized data can be deserialized and observed inside the chamber, blocking gadget-chain exploitation outside the sandbox.
Identifies and blocks malicious code introduced through deserialization of untrusted data at system boundaries.
Integrity verification of serialized information can detect tampering before deserialization occurs.
Spam protection mechanisms perform filtering and detection on inbound/outbound messages, directly compensating for missing or weak input validation of unsolicited content.
Mitigating Controls (NIST CSF 2.0) AI
Derived directly from the weakness types (CWEs) cited in the NVD entry via our AI-authored CWE→CSF cross-walk (authority under review) — links open the control.
PR.PS-02 addresses only post-deployment updates/patching and cannot prevent introduction of unsafe deserialization code, yet it can remediate some instances when the flaw exists in outdated libraries or components.
Mitigating Controls (ISO/IEC 27001:2022 Annex A) AI
Derived directly from the weakness types (CWEs) cited in the NVD entry via our AI-authored CWE→ISO cross-walk (authority under review) — links open the control.
Hardening callouts derived
DISA STIG baselines are authored against specific weaknesses. Where this CVE cites a broad weakness class, the mapping is not asserted at that abstraction (owner ruling, 2026-09-25: grade only at Base and below), so no rule is shown here rather than a rule shown against the wrong grain.
- No fine-grained STIG row for CWE-20, a class-level weakness, and none of its children carries one either.