Weekly Threat Intelligence Summary
Top 10 General Cyber Threats
Generated 2026-08-07T16:00:29.457622+00:00
- Russian State-Supported Cyber Actors Conduct Phishing Campaign Targeting Users of Zimbra Collaboration Suite (www.cisa.gov, 2026-07-21T19:08:02)
Score: 18.488
Russian State-Supported Cyber Actors Conduct Phishing Campaign Targeting Users of Zimbra Collaboration Suite Executive summary A group of Russian state-supported cyber actors has been targeting and compromising various Western government and commercial organizations using the Zimbra Collaboration Suite (ZCS) software since at least July 2025. The Russian state-supported advanced persistent threat (APT) group’s activity is tracked in the cybersecurity community under several names (see Cybersecur - Improve Router Hygiene to Protect Against Russian State-Sponsored Targeting (www.cisa.gov, 2026-07-08T18:43:49)
Score: 8.819
Russian Government-Sponsored Activity Targets Poorly Configured and Vulnerable Devices Across Critical Sectors Executive summary Russian Federal Security Service (FSB) Center 16 cyber actors continue to exploit poorly configured and vulnerable networking devices worldwide, opportunistically compromising multiple critical infrastructure sector networks. This joint Cybersecurity Advisory (CSA) builds on FBI’s Russian Government Cyber Actors Targeting Networking Devices, Critical Infrastructure Pub - Ransomware is the Scoreboard (www.recordedfuture.com, 2026-07-24T00:00:00)
Score: 8.255
Ransomware is the scoreboard for defensive architecture. Learn why traditional security methods fail and how to use AI and threat intelligence to identify and remediate critical attack paths. - Expanding AI Benchmarks in Cybersecurity Beyond Vulnerability Discovery (www.crowdstrike.com, 2026-08-06T05:00:00)
Score: 7.957 - Google’s synchronized passkeys can be stolen in ‘Pass‑ta‑key’ attacks (www.malwarebytes.com, 2026-08-05T11:11:48)
Score: 7.833
Over time, passkeys are supposed to replace passwords. But what happens when malware steals the master key? - Dealing with AI-Generated Extortion (www.recordedfuture.com, 2026-07-30T00:00:00)
Score: 7.755
Combat AI-generated extortion and fake ransomware leaks. Learn how organizations can verify data authenticity using robust governance and threat intelligence. - CrowdStrike 2026 Threat Hunting Report: Exploitation Window Closes as AI Use Accelerates (www.crowdstrike.com, 2026-08-03T05:00:00)
Score: 7.457 - Malwarebytes for Windows, now available on the Microsoft Store (www.malwarebytes.com, 2026-07-30T16:01:11)
Score: 6.867
Install Malwarebytes for Windows from the Microsoft Store with the same full protection and features. - Apple WebKit vulnerabilities reveal your IP address, despite Private Relay (www.malwarebytes.com, 2026-08-06T14:30:35)
Score: 6.023
Researchers have found three methods to bypass Apple’s Private Relay which is supposed to shield users’ IP addresses and location. - Scammers target OnlyFans users with deepfakes (www.malwarebytes.com, 2026-08-06T11:38:27)
Score: 6.003
Criminals are impersonating OnlyFans creators using AI tools in order to scam followers.
Top 10 AI / LLM-Related Threats
Generated 2026-08-07T16:00:47.617301+00:00
- Robust Context-Aware Detection of Malicious Instructions in Text (arxiv.org, 2026-08-07T04:00:00)
Score: 19.381
arXiv:2608.05430v1 Announce Type: new
Abstract: The remarkable instruction-following ability of modern LLMs has enabled their practical use as the minds of agents that can autonomously complete increasingly complex tasks. Therein, however, also lies their vulnerability to attacks which embed malicious instructions in text, common variants of which are known as indirect prompt injection (IPI). A fundamental task in addressing this vulnerability is successful segmentation of a given text into ben - PromptShield Home: Ambient Multimodal Prompt Injection Defense for Smart-Home Agents (arxiv.org, 2026-08-07T04:00:00)
Score: 18.681
arXiv:2608.05495v1 Announce Type: new
Abstract: Smart-home assistants increasingly use multimodal large language models (MLLMs) that perceive video and audio directly. This raises a safety question specific to the home: can the agent tell a genuine user command from ambient or externally-sourced content, television speech, on-screen text, or an overheard conversation, that merely looks like a command? We introduce PromptShield-Home, a pilot benchmark of realistic smart-home scenarios spanning a - MMAligner: Safeguarding Multimodal Large Language Models through Representation Calibration (arxiv.org, 2026-08-07T04:00:00)
Score: 17.681
arXiv:2608.05909v1 Announce Type: new
Abstract: Multimodal large language models (MLLMs) often refuse unsafe text prompts yet generate harmful responses to semantically equivalent multimodal inputs. Existing defenses either rely on external guardrails, which add inference overhead without repairing intrinsic flaws, or safety fine-tuning, which treats alignment as black-box optimization and may sacrifice utility or require large multimodal datasets. To identify the cause of this safety disparity - DreamGuard: Efficient Runtime Guardrail for LLM Agents via Risk-Aware World Model (arxiv.org, 2026-08-07T04:00:00)
Score: 17.681
arXiv:2608.05695v1 Announce Type: cross
Abstract: As large language model (LLM) agents increasingly invoke external tools and interact with real-world systems, unsafe actions may cause irreversible consequences on external states, user data, and downstream services. Recent runtime guardrails mitigate such risks by checking proposed actions before execution, but many remain reactive: they primarily assess the apparent safety of the current action, lacking an explicit model of how risk evolves ac - LLMs Cannot Reliably Judge (Yet?): A Comprehensive Assessment on the Robustness of LLM-as-a-Judge (arxiv.org, 2026-08-07T04:00:00)
Score: 17.681
arXiv:2506.09443v3 Announce Type: replace
Abstract: Large Language Models (LLMs) have demonstrated exceptional capabilities across diverse tasks, driving the development and widespread adoption of LLM-as-a-Judge systems for automated evaluation, including red teaming and benchmarking. However, these systems are susceptible to adversarial attacks that can manipulate evaluation outcomes, raising critical concerns about their robustness and trustworthiness. Existing evaluation methods for LLM-base - One Leak Away: How Pretrained Model Exposure Amplifies Jailbreak Risks in Finetuned LLMs (arxiv.org, 2026-08-07T04:00:00)
Score: 17.681
arXiv:2512.14751v3 Announce Type: replace
Abstract: Finetuning pretrained large language models (LLMs) has become the standard paradigm for developing downstream applications. However, its security implications remain unclear, particularly regarding whether finetuned LLMs inherit jailbreak vulnerabilities from their pretrained sources. We investigate this question in a realistic pretrain-to-finetune threat model, where an attacker has full access to a released pretrained LLM but no access to it - Text Steganography with Dynamic Codebook and Multimodal Large Language Model (arxiv.org, 2026-08-07T04:00:00)
Score: 16.681
arXiv:2604.20269v2 Announce Type: replace
Abstract: With the popularity of the large language models (LLMs), text steganography has achieved remarkable performance. However, existing methods still have some issues: (1) For the white-box paradigm, this steganography behavior is prone to exposure due to sharing the off-the-shelf language model between Alice and Bob. (2) For the black-box paradigm, these methods lack flexibility and practicality since Alice and Bob should share the fixed codebook - CREBench: Evaluating Large Language Models in Cryptographic Binary Reverse Engineering (arxiv.org, 2026-08-07T04:00:00)
Score: 14.681
arXiv:2604.03750v2 Announce Type: replace
Abstract: Reverse engineering (RE) is central to software security, particularly for cryptographic programs that handle sensitive data and are highly prone to vulnerabilities. It supports critical tasks such as vulnerability discovery and malware analysis. Despite its importance, RE remains labor-intensive and requires substantial expertise, making large language models (LLMs) a potential solution for automating the process. However, their capabilities - CARE: Pre-Execution Command Verification for Shell-Executing LLM Agents (arxiv.org, 2026-08-07T04:00:00)
Score: 14.681
arXiv:2607.21642v2 Announce Type: replace
Abstract: Large Language Model (LLM) agents are increasingly used for coding and terminal automation, making shell-command dispatch a high-stakes runtime control point. We study command-level pre-execution mediation for individual shell commands produced by LLM agents under bounded path context. Existing safeguards remain limited: generic guardrails do not model shell structure in sufficient detail, always-on LLM judges are relatively costly and variabl - Breaking Customized LLMs for Coding: Automated Red Teaming for Instruction Backdoor Attacks (arxiv.org, 2026-08-07T04:00:00)
Score: 14.381
arXiv:2608.05659v1 Announce Type: new
Abstract: LLM customization platforms allow users to build task-specific models for code intelligence tasks by embedding instructions into system prompts, without modifying the underlying model parameters. While these platforms lower the barrier to developing customized LLMs, they also introduce a new attack surface: instruction backdoor attacks, in which adversaries implant hidden malicious behaviors into customized instructions. However, existing attacks - Zero-Click AI Browser Hacking: Claude and ChatGPT Atlas Hijacked via Emails, X Posts (www.securityweek.com, 2026-08-06T12:54:09)
Score: 13.231
Zenity researchers reported the findings to Anthropic and OpenAI in late 2025 and early 2026, but they remain unpatched. The post Zero-Click AI Browser Hacking: Claude and ChatGPT Atlas Hijacked via Emails, X Posts appeared first on SecurityWeek . - Post-Hoc Trajectory-Risk Certification for Modular LLM-Based Security Agents (arxiv.org, 2026-08-07T04:00:00)
Score: 12.381
arXiv:2608.05199v1 Announce Type: new
Abstract: Autonomous security agents operate as staged pipelines, such as classifying network traffic and then attributing attacks to a specific technique. Split conformal prediction gives each stage finite-sample coverage, but deployment requires a trajectory-level guarantee across the full chain. These guarantees do not compose automatically when stages are independently trained and calibrated. Bonferroni allocation is distribution-free but conservative u - Algebraic Cryptanalytic Extraction on Hard-Label Neural Networks (arxiv.org, 2026-08-07T04:00:00)
Score: 12.381
arXiv:2608.05736v1 Announce Type: new
Abstract: Although the state-of-the-art neural network model extraction attack in the hard-label setting by Carlini et al. at EUROCRYPT 2025 has polynomial-time complexity in theory, its dual-point clustering relies on singular value decomposition (SVD) with a time complexity of $\mathcal{O}(n^2 \cdot (d^{(k)})^3)$, resulting in huge runtime in practice. To address this computational bottleneck, this work transforms Carlini et al.’s geometric-view hard - Towards a Risk Assessment of Malicious Skill Files in Coding Agents (arxiv.org, 2026-08-07T04:00:00)
Score: 12.381
arXiv:2608.05223v1 Announce Type: cross
Abstract: Autonomous coding agents are increasingly embedded in enterprise software workflows with delegated authority over connected systems. Central to this architecture is the agent skills interface: folders of instructions and scripts that agents load dynamically to specialize their behavior. This interface also widens the attack surface, letting malicious shell commands hide within natural-language skill files. We make three contributions. First, an - Behavioral Canaries: Auditing Private Retrieved Context Usage in RL Fine-Tuning (arxiv.org, 2026-08-07T04:00:00)
Score: 12.381
arXiv:2604.22191v2 Announce Type: replace
Abstract: In agentic workflows, LLMs frequently process retrieved contexts that are legally protected from further training. However, auditors currently lack a reliable way to verify if a provider has violated the terms of service by incorporating these data into post-training, especially through Reinforcement Learning (RL). While standard auditing relies on verbatim memorization and membership inference, these methods are ineffective for RL-trained mod - AI Security Leaderboard: Methodology, Results and Minimal Standard (arxiv.org, 2026-08-07T04:00:00)
Score: 12.381
arXiv:2608.03070v2 Announce Type: replace
Abstract: The AI Security Leaderboard is an independent benchmark that ranks the safeguards of frontier AI models from least to most secure. It tests models against the FAR$.$AI Minimal Standard for Safeguards, which represents a minimum bar for security: meeting it does not guarantee a secure model, but failing to meet it guarantees a lack of state-of-the-art security. Version 1.0 covers severe misuse requests across chemical, biological, radiological, - Detecting Safety Training Modification in Language Models via Activation Analysis (arxiv.org, 2026-08-07T04:00:00)
Score: 11.681
arXiv:2608.05578v1 Announce Type: new
Abstract: We introduce AMS (Activation-based Model Scanner), a tool that detects modifications to safety training in language models by measuring the geometric structure of safety-relevant concepts in activation space. Safety training creates measurable separation between harmful and benign content classes; certain safety modifications collapse or rotate this structure, while others leave it intact. We validate AMS across 14 model configurations spanning 4 - ChainClaw: A Layered Agent Framework for Reliable On-Chain Execution (arxiv.org, 2026-08-07T04:00:00)
Score: 11.681
arXiv:2608.05790v1 Announce Type: cross
Abstract: General-purpose large language model agents have achieved strong performance on tool-augmented tasks, yet they rely on assumptions break down in blockchain environments. On-chain execution is stateful, adversarial, and economically irreversible, exposing three fundamental gaps: Reactivity, Irreversibility, and Observability. We propose ChainClaw, a blockchain-native agent framework built on OpenClaw, that addresses all three gaps through a layer - Introducing Web Search on Amazon Bedrock for foundation model grounding (aws.amazon.com, 2026-08-04T18:39:14)
Score: 10.712
Today, we are introducing the general availability of Web Search on Amazon Bedrock. It is a server-side built-in tool that grounds model responses in current web knowledge. With Web Search, grounding becomes a native capability of Amazon Bedrock, with no third-party vendors to onboard, no external APIs to orchestrate, and no additional third party vendor security reviews to conduct. In this post, we walk through what Web Search on Amazon Bedrock is, why it matters, how to enable it using the Ope - Hidden prompt turns Microsoft Copilot into an AI worm (www.malwarebytes.com, 2026-07-30T12:58:45)
Score: 10.665
A new type of attack can trick Microsoft Copilot for Word into spreading hidden prompt injections from document to document. - Metasploit Framework 6.5 Released (www.rapid7.com, 2026-07-30T14:29:54)
Score: 10.58
Today we’re proud to announce that Metasploit Framework version 6.5 has been released. Over the past two years, with the help of countless contributors, we’ve added 422 new modules along with a whole slew of new features. Malleable C2 Profiles for HTTP One of the latest and most requested features is support for Malleable C2 profiles across all current Meterpreter payloads. This feature enables users to load a standard profile into Meterpreter and change the shape of its HTTP(S) traffic. All Met - LLM optimization integration for Amazon SageMaker Python SDK (aws.amazon.com, 2026-08-06T16:08:12)
Score: 10.163
The Amazon SageMaker Python SDK v3 now exposes generative AI inference recommendations in Amazon SageMaker AI directly in your notebook. Benchmark an endpoint, generate data-driven deployment recommendations, and deploy the recommended configuration without leaving your notebook workflow. - How legitimate cloud platforms enable phishers to bypass MFA (securelist.com, 2026-08-04T12:00:12)
Score: 9.746
We cover a cloud-based AitM attack scenario leveraging service workers and Ultraviolet, and provide detailed phishing hosting statistics across platforms like Cloudflare Workers, Vercel, Netlify, GitHub Pages, and IPFS. - ChainDrop: Inside a Self-Propagating npm Worm (unit42.paloaltonetworks.com, 2026-08-06T22:26:39)
Score: 9.626
Analysis of ChainDrop, an npm supply chain worm extracting GitHub Actions runner secrets and using Ethereum smart contracts for C2 routing. The post ChainDrop: Inside a Self-Propagating npm Worm appeared first on Unit 42 . - Rapid7 Analysis: Unauthenticated Remote Code Execution in JetBrains TeamCity (CVE-2026-63077) (www.rapid7.com, 2026-08-07T14:32:47)
Score: 9.485
Overview On July 27, 2026, JetBrains published a security advisory for CVE-2026-63077 , a critical unsafe deserialization vulnerability affecting JetBrains TeamCity . An attacker who can reach a TeamCity server over HTTP or HTTPS can exploit the agent polling protocol without credentials and execute operating system commands with the privileges of the TeamCity server process. JetBrains reported no known active exploitation when it disclosed the vulnerability. However, on August 5, 2026, CISA add
Auto-generated 2026-08-07
