A remote code execution vulnerability exists in Meta FAIRChem's utils load_inference_model function due to deserialization of untrusted data. Exploitation requires user interaction such as visiting a malicious page or opening a malicious file. The vulnerability has a high severity rating of 7.8 assigned by ZDI.
Meta Detectron2 contains a remote code execution vulnerability due to deserialization of untrusted data. Exploitation requires user interaction, such as visiting a malicious webpage or opening a malicious file. The vulnerability has a high severity rating of 7.8 by ZDI.
A remote code execution vulnerability exists in Meta Detectron2's load_proposals_into_dataset function. Exploitation requires user interaction, such as visiting a malicious page or opening a malicious file. The vulnerability has a high severity rating assigned by the Zero Day Initiative.
A remote code execution vulnerability exists in Meta Detectron2's CfgNode.merge_from_file function, allowing remote attackers to execute arbitrary code if a user opens a malicious file or visits a malicious page. Exploitation requires user interaction. The vulnerability has a high severity rating of 7.8 by ZDI.
A remote code execution vulnerability exists in Meta Detectron2 due to unsafe deserialization of untrusted data in LazyConfig. Exploitation requires user interaction, such as visiting a malicious webpage or opening a malicious file. The vulnerability has a high severity rating of 7.8 assigned by ZDI.
A remote code execution vulnerability exists in Meta Detectron2's DensePose component. Exploitation requires user interaction, such as visiting a malicious webpage or opening a malicious file. The vulnerability has a high severity rating of 7.8 assigned by the Zero Day Initiative.
A remote code execution vulnerability exists in Meta Detectron2's detection_checkpoint deserialization process. Exploitation requires user interaction, such as visiting a malicious webpage or opening a malicious file. The vulnerability has a high severity rating of 7.8 assigned by the Zero Day Initiative.
A remote code execution vulnerability exists in Meta fairseq's load_audio_path functionality. Exploitation requires user interaction, such as visiting a malicious webpage or opening a malicious file. The vulnerability has a high severity rating of 7.8 assigned by ZDI.
A remote code execution vulnerability exists in Meta fairseq's Wav2VecModel YAML file parsing. Exploitation requires user interaction by opening a malicious file. The vulnerability has a high severity rating of 7.8 by ZDI.
A remote code execution vulnerability exists in Meta fairseq's FilterNamesDataset component. Exploitation requires user interaction, such as visiting a malicious page or opening a malicious file. The vulnerability has a high severity rating with a CVSS score assigned by ZDI of 7.8.
A remote code execution vulnerability exists in Meta fairseq's split_by_lang component. Exploitation requires user interaction, such as visiting a malicious webpage or opening a malicious file. The vulnerability has a high severity rating of 7.8 assigned by ZDI.
A remote code execution vulnerability exists in Meta fairseq's load_speaker functionality. Exploitation requires user interaction, such as visiting a malicious webpage or opening a malicious file. The vulnerability has a high severity rating of 7.8 by ZDI.
A remote code execution vulnerability exists in Meta fairseq's VGGTransformerModel YAML file parsing. Exploitation requires user interaction by opening a malicious file. The vulnerability has a high severity rating of 7.8 by ZDI.
A remote code execution vulnerability exists in Meta fairseq's FairseqAdafactor YAML file parsing. Exploitation requires user interaction by opening a malicious file. The vulnerability has a high severity rating of 7.8 by ZDI.
Able2Shell is a collection of 33 reported zero-day vulnerabilities in N-able N-central, disclosed publicly with proof-of-concept code. Among these, seven critical vulnerabilities have been officially recognized and patched by N-able in the N-central 2026.4 Hotfix 2 release. The vulnerabilities include high-severity issues such as XML external entity (XXE) attacks, server-side request forgery (SSRF), OGNL injection, local privilege escalation, remote code execution, and stored cross-site scripting (XSS). One exploit chain allows unauthenticated remote command execution as the nable user, and others can lead to root privileges through chained exploits. N-able has confirmed and patched the most critical issues but has not mapped all 33 findings individually. The affected versions are N-central prior to 2026.4.1.26. No confirmed customer compromises have been reported as of November 10, 2026.
Medium6.5KEVEPSS 1.6%Krebs on SecurityCVE-2026-124194
Microsoft released a large Patch Tuesday update addressing 1,093 vulnerabilities, including 63 critical flaws affecting Windows and related software. Among these is a zero-day vulnerability (CVE-2026-124194) in Microsoft Outlook that is actively exploited. Nearly half of the critical bugs allow remote code execution, and many others enable privilege escalation. Other major vendors including Apple, Google, Mozilla, and Adobe also issued security updates this week. Microsoft credits AI advancements for the increased volume of identified and fixed vulnerabilities. The update is part of a continuous effort to mitigate risks from vulnerabilities actively exploited or potentially exploitable in the wild.
A high-severity use-after-free vulnerability (CVE-2026-125350) in the XML component of Google Chrome was reported anonymously on November 6, 2026, and is already exploited in the wild. This zero-day is one of eight exploited Chrome zero-days patched in 2026. The vulnerability is addressed in Chrome 157.0.8103.50/.51 for Windows and Mac, and 157.0.8103.50 for Linux. Google has not disclosed technical details or attack specifics. Additional high- and medium-severity vulnerabilities were also fixed in this update.
A remote code execution vulnerability exists in Wireshark due to an integer overflow in SMB Object Export processing. Exploitation requires user interaction. The vulnerability has a high severity rating of 8.8 assigned by ZDI.
A remote code execution vulnerability exists in NVIDIA Megatron-LM due to unsafe deserialization of untrusted checkpoint data. Exploitation requires user interaction, specifically opening a malicious file. The vulnerability is assigned CVE-2026-65110 with a high severity rating by ZDI.