CVE-2026-5713
Information Disclosure via Profiling and Asyncio in Python
Publication date: 2026-04-14
Last updated on: 2026-04-15
Assigner: Python Software Foundation
Description
Description
CVSS Scores
EPSS Scores
| Probability: | |
| Percentile: |
Meta Information
Affected Vendors & Products
| Vendor | Product | Version / Range |
|---|---|---|
| python | cpython | 3.15 |
| python | cpython | 3.14 |
| python | python | 3.15 |
| python | python | 3.14 |
Helpful Resources
Exploitability
| CWE ID | Description |
|---|---|
| CWE-125 | The product reads data past the end, or before the beginning, of the intended buffer. |
| CWE-121 | A stack-based buffer overflow condition is a condition where the buffer being overwritten is allocated on the stack (i.e., is a local variable or, rarely, a parameter to a function). |
Attack-Flow Graph
AI Powered Q&A
Can you explain this vulnerability to me?
CVE-2026-5713 is a vulnerability in Python's remote debugging feature affecting the "profiling.sampling" module (Python 3.15+) and asyncio introspection capabilities (3.14+). It allows a malicious or compromised Python process to exploit the remote debugging connection to read and write memory addresses in a privileged process.
The root cause is insufficient validation of debug offset tables used during remote debugging. These tables, which contain sizes and offsets for unwinding and introspection, were previously trusted without strict checks. Malformed or maliciously crafted tables can cause out-of-bounds memory accesses or undefined behavior.
Exploitation requires persistently and repeatedly connecting to the target process, as the connecting process is likely to crash frequently due to Address Space Layout Randomization (ASLR). The vulnerability is mitigated by introducing comprehensive validation of these offset tables before use, rejecting malformed or inconsistent data upfront.
How can this vulnerability impact me? :
This vulnerability can allow an attacker to perform unauthorized out-of-bounds read and write operations in a privileged Python process during remote debugging sessions.
Such unauthorized memory access could lead to information disclosure, corruption of memory, or potentially arbitrary code execution depending on the context and privileges of the targeted process.
However, exploitation is difficult as it requires persistent and repeated connections to the target process, and the connecting process is likely to crash frequently due to ASLR protections.
How can this vulnerability be detected on my network or system? Can you suggest some commands?
The vulnerability involves the Python remote debugging feature, specifically the "profiling.sampling" module and asyncio introspection capabilities. Detection would involve monitoring for processes using remote debugging connections, especially those repeatedly connecting to privileged Python processes.
You can use the asyncio introspection commands introduced in Python 3.14 and later to inspect running asyncio processes:
- python -m asyncio ps
- python -m asyncio pstree
These commands help identify asyncio tasks and their state, which may assist in detecting suspicious remote debugging activity. However, no specific detection commands for the vulnerability itself are provided.
What immediate steps should I take to mitigate this vulnerability?
Immediate mitigation involves applying the security patches that introduce strict validation of remote debug offset tables before they are used by the unwinder during remote debugging.
The fix validates externally read debug-offset metadata and rejects malformed or inconsistent offset and size tables upfront, preventing unsafe memory accesses.
Ensure your Python installation is updated to a version that includes the patches merged in pull requests such as gh-148178, gh-148179, and gh-148187, which harden the remote debugging subsystem.
Additionally, avoid allowing untrusted or potentially malicious Python processes to connect via remote debugging to privileged processes.
How does this vulnerability affect compliance with common standards and regulations (like GDPR, HIPAA)?:
The provided information does not specify any direct impact of CVE-2026-5713 on compliance with common standards and regulations such as GDPR or HIPAA.