A vulnerability in Cisco NX-OS Software could allow an unauthenticated, remote attacker to exhaust system resources, causing a denial of service (DoS) condition.
This vulnerability exists because rate limiting was improperly applied to some protocols. An attacker could exploit this vulnerability by sending a high rate of UDP or TCP connections to a data plane interface on an affected device. A successful exploit could allow the attacker to cause instability to various routing and control plane protocols through some packet loss and temporary disruptions, causing a DoS condition. This DoS condition will clear without manual intervention soon after the high rate of traffic is stopped.
CVE-2026-20328 is a critical vulnerability in the web-based management interface of Cisco License On-Prem, formerly known as Cisco Smart Software Manager On-Prem. It arises from improper authorization checks during the password reset process. An unauthenticated remote attacker can exploit this flaw by sending a crafted request to reset the password of any account, including high-privileged administrative users. This could allow unauthorized access to the application with elevated privileges.
CVE-2026-20032 is a medium severity vulnerability in the Python interpreter of Cisco NX-OS Software. It allows an authenticated local attacker with low privileges to escape the Python sandbox and gain unauthorized access to the underlying operating system. The issue arises from insufficient validation of user-supplied input, enabling manipulation of specific Python interpreter functions. Exploitation could lead to arbitrary command execution with the privileges of the authenticated user.
A vulnerability in the endpoint group (EPG) contract functionality of Cisco Nexus 9000 Series Fabric Switches in ACI Mode could allow an unauthenticated, remote attacker to bypass configured EPG contracts.
This vulnerability is due to an improper control with EPG contracts. An attacker could exploit this vulnerability by sending IPv4 or IPv6 packets using UDP source and destination ports that are assigned to DHCP traffic through an affected device. A successful exploit could allow the attacker to bypass EPG contracts on the affected device.
CVE-2026-20321 is a vulnerability in the web-based management API of Cisco Application Policy Infrastructure Controller (APIC) that allows an authenticated attacker with administrative credentials to execute arbitrary commands as root. The flaw arises from insufficient input validation of user-controlled command arguments. Exploitation requires valid admin credentials and could lead to full system compromise with root privileges. The CVSS score is 6.5 (medium severity). No patch or remediation information is provided in the available data.
Homer, an open source telecom observability software, had a critical vulnerability (CVE-2026-62253) prior to version 11.0.283 where its JWT middleware functions allowed unauthenticated access to protected API endpoints. This was due to the JWT secret defaulting to an empty string, causing the middleware to bypass authentication checks. The issue was fixed in version 11.0.283.
CVE-2026-62251 is a high-severity SQL injection vulnerability in sipcapture homer, an open source telecom observability software. Versions prior to 11.0.283 allow authenticated users to execute arbitrary SQL commands via the V4StatisticsQuery handler, which fails to validate user-supplied SQL queries properly. This issue is patched in version 11.0.283.
PraisonAI is a multi-agent teams system. Prior to version 4.6.78, the `deploy/api.py` module generates Python server code by directly interpolating the `agents_file` parameter into an f-string that is then written to a file and executed via `subprocess.Popen()`. An attacker who controls the `agents_file` value (via CLI argument, configuration, or upstream API) can inject arbitrary Python code. Version 4.6.78 patches the issue.
CVE-2026-107166 is a medium severity vulnerability in Open5GS up to version 2.7.7. It affects the GTP-U Receive Path component, specifically the ogs_pfcp_xact_local_create function in src/upf/gtp-path.c. The flaw allows remote attackers to cause allocation of resources, potentially impacting system stability. A patch has been released to address this issue.
Microsoft's FORGE Lab is advancing AI-native vulnerability research, discovering and helping remediate numerous vulnerabilities across Windows and open-source projects including the Linux kernel. Their work from May to September 2026 led to 140 CVEs assigned for Windows and 155 validated reports for open-source projects. The research highlights challenges beyond discovery, emphasizing the need for scalable validation and remediation processes to keep pace with AI-generated findings. Key lessons include shifting focus from capability to scale, optimizing reasoning economics, and coordinating validation and remediation efforts. This approach aims to sustain vulnerability research at scale by integrating automated validation tools and human coordination.
CVE-2026-92414 is a critical session fixation and session reuse vulnerability in Apache Jackrabbit's WebDAV server. The flaw allows the server to attach a cached authenticated session based solely on matching certain header tokens (Lock-Token, TransactionId, SubscriptionId, If-header) without verifying credentials. This can lead to session reuse across different users. Versions 2.20.0 through 2.20.17, 2.22.0 through 2.22.4, and 2.23.0 through 2.23.5 are affected. Fixed versions are 2.20.18, 2.22.5, and 2.23.6.
Microsoft will block .msix and .msixbundle file attachments in Outlook Web and the new Outlook Windows client starting in November 2026. These file types, used for modern Windows installation packages, will be added to the default blocked attachments list in Exchange Online's OWA Mailbox policies. After the update, users will be unable to send, receive, open, or download these attachments by default. Administrators can whitelist these file types if necessary. This change is part of Microsoft's ongoing efforts to enhance security by blocking potentially unsafe file attachments that have been abused in attacks.
PoeLLM is malware involved in a cryptomining campaign targeting exposed AI servers. The malware infects these servers to convert them into scanners and exploit launchpads for further attacks. There is no detailed technical information or confirmed exploits in the wild provided.
A flaw was found in source-to-image. When unpacking archive files, the application fails to properly sanitize symbolic links pointing to absolute file paths. An attacker who supplies a malicious builder image can exploit this vulnerability by embedding links pointing outside the extraction directory. This allows the attacker to bypass sandbox boundaries, potentially leading to unauthorized information disclosure or file modification on the host system.
CVE-2026-107169 is a vulnerability in the m17n-lib component of Red Hat Enterprise Linux 10. It involves a null pointer dereference triggered by specially crafted or truncated UTF-8 input during text processing. This flaw causes the affected application to crash unexpectedly, resulting in a denial of service (DoS). The vulnerability has a medium severity rating with a CVSS score of 6.2. No affected versions or patch information are explicitly provided in the available data.