OpenSSL critical buffer overflow CVE-2026-121481
Archived — this story has rotated out of today’s deck. It is kept here in full.
The gist
OpenSSL fixed a heap buffer overflow in EVP_EncodeBlock() in versions 4.1.1 and 4.0.4.
The 9.8 severity score is disputed, and no public exploits are known.
Background
OpenSSL is the encryption library named in the CVE record, and the flaw sits in EVP_EncodeBlock(), a function that base64-encodes data. The CVE record says calling it with an input length close to INT_MAX can overflow the encoded output size calculation and write past the end of the caller-supplied output buffer. It is drawing attention because OpenSSL sits on nearly every server, desktop and phone, per LinkedIn poster Omar Ahmed, and because a 9.8 severity score circulated quickly while the OpenSSL pull request discussion questioned whether the issue deserved a CVE at all.
How it unfolded
- Nov 18, 2026CVE-2026-121481 is published, listing affected versions as 4.1.0 before 4.1.1 and 4.0.0 before 4.0.4; the record describes the overflow in EVP_EncodeBlock()'s output size calculation.
- Nov 18, 2026OpenCVE lists the vulnerability with no CVSS score and no EPSS data; Strix lists the same affected ranges with NVD analysis pending and severity unknown.
- Nov 19, 2026cve.now posts a CRITICAL 9.8 rating, and Radoslav Krehlik reports the same score, saying no public exploits are known but users should update promptly.
- Nov 19, 2026Omar Ahmed reports the CISA SSVC assessment: exploitation none, automatable yes, technical impact total; the CVE is not in the CISA KEV catalog.
- Nov 20, 2026Omar Ahmed states the documented route is the openssl s_client command line via a very large -proxy_pass value, including one supplied through an environment variable.
- Nov 20, 2026Myria posts questioning what application encodes 2 GiB of data in base64, disputing the 9.8 score.
Key terms
- heap buffer overflow
- A software bug where a program writes more data into a memory buffer than it can hold, overwriting nearby memory.
- CVE
- A unique identifier in the CVE system for a publicly known information-security vulnerability.
- CVSS
- A framework for rating the severity of computer system vulnerabilities on a scale from 0 to 10.
Source: Wikipedia
Who’s saying what
- Caution
- Radoslav Krehlik rates the flaw CVSS 9.8 and urges prompt updating, noting no public exploits are known.
- Analysts
- Omar Ahmed notes the advisories don't say whether default setups are exposed, arguing unknown is not the same as safe and that openssl version proves nothing because distributions backport fixes under older version numbers.
- Expert
- OpenCVE says the overflow might enable arbitrary code execution if an attacker controls the input, but this is inferred from the memory corruption potential.
- Public
- Myria disputes the 9.8 score, asking what application uses OpenSSL to encode 2 GiB of data in base64.
Still unverified
Whether default setups are exposed is not stated in the advisories, per Omar Ahmed; arbitrary code execution is inferred from memory corruption potential, not confirmed; severity scoring is pending NVD analysis, with Strix listing severity as unknown; an OpenSSL pull request reviewer said the report's reproducer is completely synthetic and probably should not be a CVE without evidence it really happens.