What changed in AI security, Oct 26 to Nov 1, 2026
Oct 26 to Nov 1, 2026 (ISO week 2026-W44). Weeks run Monday to Sunday in UTC.
378 records published, +43 on the previous week: 91 vulnerabilities (+40), 0 incidents (no change), 96 research items (-28), 189 news items (+32), 2 policy items (-1).
Critical and high advisories
Vulnerability records rated critical or high, newest first. Showing 25 of 57.- Critical
Beta9 is vulnerable to a Container Escape via Unvalidated Sandbox File Paths
JFrog Security Research (Vulnerabilities) - Critical
Prefab UI is vulnerable to Remote Code Execution via a Pyodide js Bridge to Deno
JFrog Security Research (Vulnerabilities) - Critical
Beta9 is vulnerable to Host Filesystem Exposure via Volume Path Traversal
JFrog Security Research (Vulnerabilities) - Critical
Beta9 is vulnerable to a Build Worker Sandbox Escape via Chroot-only Buildah Isolation
JFrog Security Research (Vulnerabilities) - High
CVE-2026-63378: TAKO AI Agent for Okta code execution via execute-script endpoint
CVE-2026-63378NVD/CVE Database - Critical
CVE-2026-63377: TAKO AI Agent for Okta unauthenticated admin account creation via setup endpoint
CVE-2026-63377NVD/CVE Database - High
CVE-2026-119337: RAGFlow agent chat completions API bypasses access check on private agents
CVE-2026-119337NVD/CVE Database - High
CVE-2026-62331: praktor path traversal via Telegram document upload filenames
CVE-2026-62331NVD/CVE Database - High
CVE-2026-119978: Flowise sandbox escape to OS command execution via node-custom-function endpoint
CVE-2026-119978NVD/CVE Database - High
CVE-2026-61789: DBHub read-only SQL validation bypass on MySQL and MariaDB connectors
CVE-2026-61789NVD/CVE Database - Critical
CVE-2026-61730: SOCFortress CoPilot unauthenticated access to ScoutSuite reports and GCP keys
CVE-2026-61730NVD/CVE Database - Critical
CVE-2026-61729: SOCFortress CoPilot webhook authentication bypass via default header secret
CVE-2026-61729NVD/CVE Database - High
CVE-2026-61708: SOCFortress CoPilot data store upload allows cross-tenant object overwrite
CVE-2026-61708NVD/CVE Database - High
CVE-2026-119408: n8n remote code execution via deprecated Function and LangChain Code nodes
CVE-2026-119408NVD/CVE Database - Critical
CVE-2026-61707: SOCFortress CoPilot focuses on providing a single pane of glass for all your security operations needs. Prior to…
CVE-2026-61707NVD/CVE Database - High
CVE-2026-118761: Kubeflow Pipelines decompression bomb in pipeline archive upload
CVE-2026-118761NVD/CVE Database - High
CVE-2026-118759: Kubeflow Pipelines run log endpoint exposes logs without authorization
CVE-2026-118759NVD/CVE Database - High
CVE-2026-118758: Kubeflow Pipelines path traversal in GET /artifacts/get volume handler
CVE-2026-118758NVD/CVE Database - High
GHSA-w9x9-j56r-h98r: @eeacms/volto-chatbot has Unauthenticated Danswer API Confused Deputy Proxy
GitHub Advisory Database - High
GHSA-q344-5v34-gm84: @sylphx/pdf-reader-mcp: HTTP transport accepts unauthenticated requests despite a configured MCP_API_KEY
CVE-2026-62264GitHub Advisory Database - High
GHSA-26mh-57q7-jfvr: PraisonAI vulnerable to Remote Code Execution via Broken AST Sandbox
CVE-2026-61438GitHub Advisory Database - High
CVE-2026-119120: Crawl4AI Docker API lets clients choose the HuggingFace model it loads
CVE-2026-119120NVD/CVE Database - Critical
CVE-2026-110186: IBM Langflow OSS remote arbitrary code execution via code generation
CVE-2026-110186NVD/CVE Database - Critical
CVE-2026-107329: IBM Langflow OSS arbitrary code execution via code generation
CVE-2026-107329NVD/CVE Database - Critical
CVE-2026-107328: IBM Langflow OSS remote code execution through eval/exec
CVE-2026-107328NVD/CVE Database
Exploitation signals
Vulnerabilities published in the week that are listed in the CISA Known Exploited Vulnerabilities catalog or have an EPSS score of 10% or more.No vulnerability published in this week is listed as exploited or has an EPSS score of 10% or more.
Packages that began delegating to a language model
Exposure Registry packages whose first release declaring an LLM SDK, agent framework or MCP dependency was published in the week.No tracked package published its first release with an LLM SDK, agent framework or MCP dependency in this week.
Topics that moved
Largest increases over the mean of the 4 previous weeks, for topics with at least 3 records in the week.| Topic | Records | Weekly mean, previous 4 | Difference |
|---|---|---|---|
| Model Context Protocol | 20 | 11.8 | +8.3 |
| Prompt injection and jailbreaks | 19 | 11.8 | +7.3 |
| AI agents | 44 | 40.0 | +4.0 |
| Coding assistants | 5 | 2.0 | +3.0 |
| Model and package supply chain | 3 | 2.0 | +1.0 |
Research
Peer-reviewed first, then newest. Showing 8 of 96.Critiquing AI Surveys with a Case Study on Temporal Features for Trustworthy Computer Vision
Peer-reviewedACM Digital Library (TOPS, DTRAP, CSUR)Vulnerabilities and Defenses in Audio-Visual Attacks: A Survey From Audio to Multimodal Models
Peer-reviewedACM Digital Library (TOPS, DTRAP, CSUR)Evaluating security and inclusivity in LLM-generated code
Peer-reviewedElsevier Security JournalsMedBackdoorGuard: Detecting and Mitigating Backdoor Attacks in Deep Learning Models for MRI-Based Alzheimer’s Disease and Mild Cognitive Impairment Diagnosis
Peer-reviewedOpenAlex (peer-reviewed AI security)Beyond public access in LLM pre-training data
Peer-reviewedOpenAlex (peer-reviewed AI security)Design of an Iterative Adversarial Resilient Graph Neural Networks for Intrusion Detection in Software Defined Networks and Deployments
Peer-reviewedOpenAlex (peer-reviewed AI security)Model Extraction of Convolutional Neural Networks with Max-Pooling
Peer-reviewedOpenAlex (peer-reviewed AI security)FairLend: Privacy-preserving gender-fairness auditing for loan processing
Peer-reviewedElsevier Security Journals
Policy and regulation
Newest first.Generated from the AI Sec Watch database at . Every item links to its record.