Weekly Threat Report 2026-11-16
Weekly Threat Intelligence Summary
Top 10 General Cyber Threats
Generated 2026-11-16T05:00:07.031688+00:00
- #StopRansomware: SafePay Ransomware (www.cisa.gov, 2026-10-19T19:18:07)
Score: 14.233
Advisory at a Glance Title #StopRansomware: SafePay Ransomware Original Publication October 20, 2026 Executive Summary This advisory details recent activity by the SafePay ransomware group. SafePay actors have been highly active since August 2024, targeting businesses and critical infrastructure across North America, South America, Europe, and Australia. SafePay actors typically gain access to networks by stealing or guessing passwords, exploiting vulnerabilities in security systems, and using p - Update Chrome now to fix a flaw attackers are exploiting (www.malwarebytes.com, 2026-11-10T13:23:00)
Score: 12.258
Google has released Chrome updates for desktop and Android to patch a zero-day vulnerability already being exploited in attacks. - GhostFrame: A new ESXi Implant Linked to Settra Ransomware (www.recordedfuture.com, 2026-11-10T00:00:00)
Score: 10.165
GhostFrame is a new post-exploitation framework targeting VMware ESXi environments, linked to TAG-218 and Settra ransomware. - November 2026 Patch Tuesday: 1,252 CVEs Include One Zero-Day and 68 Critical Vulnerabilities (www.crowdstrike.com, 2026-11-10T06:00:00)
Score: 9.707 - Microsoft Patch Tuesday fixes Outlook flaw already under attack (www.malwarebytes.com, 2026-11-11T11:25:05)
Score: 9.411
Microsoft has fixed an actively exploited Outlook vulnerability. Check for Office updates separately. - Remus malware claims Gmail access even after you change your password (www.malwarebytes.com, 2026-11-11T13:42:10)
Score: 7.427
Developers of the infostealer Remus claim their malware can retain access to a victim’s Gmail even after a password reset. - How Adversaries Are Turning the Vulnerability Ecosystem Against Defenders (www.crowdstrike.com, 2026-11-05T06:00:00)
Score: 6.374 - October 2026 CVE Landscape (www.recordedfuture.com, 2026-11-06T00:00:00)
Score: 6.299
In October 2026, Insikt Group® identified 82 high-impact vulnerabilities that should be prioritized for remediation, 52 of which had a Very Critical Recorded Future Risk Score. This represents a 13% decrease from last month. - New York City votes to ban surveillance pricing (www.malwarebytes.com, 2026-11-13T16:52:36)
Score: 5.782
The bill would stop companies from using personal data to set individualized prices. Here’s what that means for shoppers. - Google pushes second Chrome update this week (www.malwarebytes.com, 2026-11-13T11:20:50)
Score: 5.744
Google has pushed out a second Chrome update this week, fixing security flaws that could allow attackers to run malicious code.
Top 10 AI / LLM-Related Threats
Generated 2026-11-16T06:00:20.850904+00:00
- Forget the Lineage: Forward-Secure, Lineage-Complete Cryptographic Erasure for LLM-Agent Memory (arxiv.org, 2026-11-16T05:00:00)
Score: 19.79
arXiv:2611.18133v1 Announce Type: new
Abstract: Large language model (LLM) agents keep a long-term memory that the model itself writes: summaries, extracted facts and embeddings derived from what users said. Deleting a message is therefore not enough: its derivatives, storage residue and backups survive. We define lineage-complete forward-secure erasure: after a forget request, an adversary that later obtains all storage and the agent's current keys learns nothing about the forgotten input - I2C-GRPO: Insight Importance Checklist for Reinforcement Learning in a Multi-Agent DDoS Alert Triage Workflow (arxiv.org, 2026-11-16T05:00:00)
Score: 17.79
arXiv:2611.17534v1 Announce Type: new
Abstract: Large language model (LLM) agents are increasingly used for investigative tasks over incomplete evidence, such as cybersecurity alert triage, but reinforcement learning (RL) for such tasks is difficult: intermediate steps have no verifiable labels, rollouts are costly, and outcome rewards are sparse. We propose Insight Importance Checklist (I2C)-GRPO, a checklist reward for intermediate investigation steps. For each decision point, a panel of expe - LLMs are Vulnerable to Malicious Requests with Indistinguishable Intent (arxiv.org, 2026-11-16T05:00:00)
Score: 17.79
arXiv:2611.17818v1 Announce Type: new
Abstract: Jailbreak attacks can circumvent the safety mechanisms of Large Language Models (LLMs) and lead to the generation of harmful or policy-violating content. We first examine eight jailbreak methods against an LLM-based filter and observe that requests resembling normal user queries tend to have higher filter pass rates than those containing obvious anomalous patterns. This observation motivates the hypothesis that resemblance to normal requests may c - Cyber-Zero: Training Cybersecurity Agents without Runtime (arxiv.org, 2026-11-16T05:00:00)
Score: 17.79
arXiv:2508.00910v3 Announce Type: replace
Abstract: Large Language Models (LLMs) have achieved remarkable success in software engineering tasks when trained with executable runtime environments, particularly in resolving GitHub issues. However, such runtime environments are often unavailable in other domains, especially cybersecurity, where challenge configurations and execution contexts are ephemeral or restricted. We present Cyber-Zero, the first runtime-free framework for synthesizing high-q - AttentionSwitch: Jailbreak Defense for LLMs via Joint Learning of Sparse Attention-Head Selection and Scaling (arxiv.org, 2026-11-16T05:00:00)
Score: 17.79
arXiv:2611.08752v2 Announce Type: replace-cross
Abstract: Large language models have achieved remarkable performance across a wide range of tasks, yet they remain susceptible to jailbreak attacks that circumvent safety mechanisms and elicit harmful or unintended outputs. Although safety can be improved through model fine-tuning, such approaches can be computationally expensive and may degrade previously acquired capabilities through catastrophic forgetting. Recent jailbreak defenses have theref - Can LLMs Write Backdoors? An Empirical Security Study of Weight-Space Programming (arxiv.org, 2026-11-16T05:00:00)
Score: 14.79
arXiv:2611.17395v1 Announce Type: new
Abstract: Recent advances in weight-space programming (WSP) have demonstrated that large language models (LLMs) can directly inspect and modify neural network weights without conventional gradient-based training. While this capability opens new possibilities for model improvement, it also raises an important security question: Can LLMs exploit weight-space programming to inject backdoors into neural networks? To answer this question, we present an empirical - Condensing Law of Security Information (arxiv.org, 2026-11-16T05:00:00)
Score: 14.79
arXiv:2611.17918v1 Announce Type: new
Abstract: Scaling laws indicate that the generalization error of a large language model (LLM) decreases monotonically with model size once data is sufficient, and the corollary that larger models outperform smaller ones is now widely accepted. We argue that this corollary fails in security: decisions there hinge on anomalous inputs, which are long-tailed (e.g., zero-day attacks) and therefore poorly served by likelihood fitting. Worse still, the capability - From Bytes to Behavior: Request-Level Attribution from Live GPU Hidden States (arxiv.org, 2026-11-16T05:00:00)
Score: 14.79
arXiv:2611.18272v1 Announce Type: new
Abstract: Existing memory forensics for machine learning systems typically targets static artifacts, such as model weights or prompt logs, rather than the transient computational states of a live inference process. This gap limits an examiner's ability to analyze the specific behavior of a running large language model (LLM) service that cannot be restarted or modified. We present a method for acquiring per-layer residual-stream activations from a live - Training Language Model Agents to Find Vulnerabilities with CTF-Dojo (arxiv.org, 2026-11-16T05:00:00)
Score: 14.79
arXiv:2508.18370v3 Announce Type: replace-cross
Abstract: Large language models (LLMs) have demonstrated exceptional capabilities when trained within executable runtime environments, notably excelling at software engineering tasks through verified feedback loops. Yet, scalable and generalizable execution-grounded environments remain scarce, limiting progress in training more capable ML agents. We introduce CTF-Dojo, the first large-scale executable runtime tailored for training LLMs with verifi - REFORGE: A Method for Benchmarking LLMs' Reverse Engineering Capabilities in Decompiled Binary Function Naming (arxiv.org, 2026-11-16T05:00:00)
Score: 14.79
arXiv:2607.07738v3 Announce Type: replace-cross
Abstract: Large language models (LLMs) are increasingly applied to reverse-engineering tasks, and recent threat-intelligence reporting shows them operating inside live offensive-security workflows. Claims about their capability, however, outpace our ability to measure it. Existing benchmarks for LLM-assisted binary analysis treat the construction of function-level ground truth as a solved pre-processing step and report accuracy without disclosing - Route coding agent requests to the right model on Amazon Bedrock with NVIDIA NeMo Switchyard (aws.amazon.com, 2026-11-13T22:30:12)
Score: 13.849
Coding agents work in tasks, and the number that matters is cost per solved task, not price per token. In this post, you use the AWS CDK to deploy LiteLLM with the NVIDIA NeMo Switchyard routing plugin on Amazon Bedrock, then measure cost per solved task for a routed model group against single-model baselines. - Building a self-service LLM gateway with Amazon Bedrock and LiteLLM (aws.amazon.com, 2026-11-12T18:41:12)
Score: 13.573
Build a self-service LLM gateway with LiteLLM on AWS Fargate that puts Amazon Bedrock foundation models behind an OpenAI-compatible API. Developers get instant access through a single API key, while platform teams gain per-user cost controls, quota isolation, and observability. - The November 2026 Security Update Review (www.thezdi.com, 2026-11-10T19:02:41)
Score: 12.601
The penultimate month of 2026 bring with it the usual storm of patches from Microsoft and Adobe. I’ve been prototyping some new ways to display this data, but as always, I’m open to suggestions. Seriously – hit me up on Bluesky if you have any ideas.If you’d rather watch the full video recap covering the entire release, you can check it out here: Adobe Patches for November 2026 For the first part of the November release, Adobe released nine bulletins addressing 110 unique CVEs in Adobe Connect, - CANTraj: Recovering Vehicle Trajectories from CAN Signals in Naturalistic Driving (arxiv.org, 2026-11-16T05:00:00)
Score: 12.49
arXiv:2611.18442v1 Announce Type: new
Abstract: Open driving releases and usage-based insurance programs often claim to prevent privacy breaches by omitting GPS records when sharing or collecting vehicle data. However, removing GPS coordinates from driving logs does not erase the location information carried by vehicle signals. While prior route-inference attacks have demonstrated that speed, steering, and driving events can be matched to road networks to reconstruct trajectories, existing stud - Policy and Disclosure: 2026 Edition (projectzero.google, 2026-11-13T08:00:00)
Score: 11.805
Since 2021, Project Zero has followed a “90 + 30” vulnerability disclosure policy, allowing vendors up to 90 days to remediate a vulnerability plus an extra 30 days “soak time” for patch rollout. Our goal was to encourage faster patching and give vendors time to deliver patches to users. The advent of AI-assisted vulnerability discovery and exploitation has changed the threat landscape such that the vulnerability remediation timeline encouraged by our disclosure policy is clearly insufficient. L - Beyond Attack-Based Accuracy: Rethinking Adversarial Robustness Evaluation Through Structural Gradient Alignment (arxiv.org, 2026-11-16T05:00:00)
Score: 11.79
arXiv:2611.17377v1 Announce Type: new
Abstract: Visual adversarial robustness remains a fundamental security property as vision systems evolve from conventional classifiers to vision-language models (VLMs) and large vision-language models (LVLMs), with visual inputs remaining an attack surface across generations. Yet existing evaluation methods often depend on prescribed attacks, in-distribution data, extensive probing, or classifier-specific formulations, limiting external auditing of diverse - How Natera designed an agentic AI system for lab ops with Amazon Bedrock AgentCore (aws.amazon.com, 2026-11-12T21:47:07)
Score: 11.604
Natera built SWARM, an agentic AI system on Amazon Bedrock AgentCore that captures institutional lab-operations knowledge and deploys it at scale. This post walks through the full architecture, from knowledge base construction to deployment and evaluation, delivering about 130x faster first response and 96.2% automated coverage across support tickets. - Filter generation in practice: from algorithms to LLMs (securelist.com, 2026-10-30T10:00:44)
Score: 11.492
We automate filter generation for security alerts using an algorithm, a large language model, and a group of LLM agents. - Rapid7 Achieves ISO/IEC 42001 Certification for its AI Management System (www.rapid7.com, 2026-11-04T16:25:55)
Score: 11.046
Exciting news for AI governance! Rapid7 has achieved ISO/IEC 42001:2023 certification for our Artificial Intelligence Management System (AIMS). The certificate was issued by Schellman Compliance, LLC, an ANAB accredited certification body, following an independent audit of how we govern the AI systems we develop in our role as an AI provider. What is ISO/IEC 42001? ISO/IEC 42001 is the first international management system standard for artificial intelligence. Where ISO/IEC 27001 focuses on info - OpenAI Codex on Amazon Bedrock (aws.amazon.com, 2026-11-11T23:35:47)
Score: 10.384
A practical guide for enterprises rolling out OpenAI Codex to developers on AWS, with Amazon Bedrock as the model backend, covering the access, governance, and cost decisions that platform, security, and finance teams need to align on. - LLMON: An LLM-Native Markup Language, now in Granite 4.1 3B and 8B (huggingface.co, 2026-11-12T19:07:35)
Score: 9.778 - From prototype to production: Engineering agents with Agent DLC (aws.amazon.com, 2026-11-12T15:38:47)
Score: 9.543
Agents stall between demo and release when teams lack a repeatable way to measure how good an agent is and whether the last change helped. This post covers three agent properties that break traditional CI/CD, a measured case where an aggregate metric hides a coverage gap, and Agent DLC: one lifecycle, two loops, and one explicit standard for judging every change. - Evolutionary-based Transfer Learning for Physical Attack on End-to-End Autoencoder Communications (arxiv.org, 2026-11-16T05:00:00)
Score: 9.49
arXiv:2611.17393v1 Announce Type: new
Abstract: End-to-end autoencoder communication has emerged as a promising means of achieving efficient data transmission across various domains. Nevertheless, vulnerabilities have recently been revealed in these systems by adversarial attacks, wherein malevolent entities manipulate the transmitted signals to deceive the autoencoder. This paper leverages the applications of evolutionary algorithms to solve a challenging optimization problem of launching phys - Evaluating the Robustness of PRNU Source-Camera Attribution to the Fingerprint-Copy Attack: An Open Toolkit with Triangle-Test Detection (arxiv.org, 2026-11-16T05:00:00)
Score: 9.49
arXiv:2611.17747v1 Announce Type: new
Abstract: Photo-Response Non-Uniformity (PRNU) is the de facto standard trace for sensor-level source-camera attribution, and a high Peak-to-Correlation-Energy (PCE) between a questioned image and a camera reference fingerprint is routinely treated as strong evidence of common origin. The fingerprint-copy attack of Goljan, Fridrich and Chen (2011) shows that this signal can be forged: an adversary who estimates a target camera's fingerprint from that c - Securing Multi-Agent Systems by Design (arxiv.org, 2026-11-16T05:00:00)
Score: 9.49
arXiv:2611.17883v1 Announce Type: new
Abstract: Multi-agent systems solve complex tasks by decomposing them into subtasks, delegating the subtasks to agents, and aggregating the agents' responses using an aggregation rule. Malicious agents can manipulate the final solution by providing malicious responses to their assigned subtasks. In this paper, we first propose response manipulation attacks (ReMA), which iteratively optimize malicious responses to maximize attack effectiveness. Specific
Auto-generated 2026-11-16