Off the Wire
What’s New
New and updated since 31 August 2026. The latest stories filed to the Digest, newest first.
New since last issue
526 stories- Attackers conceal phishing lures using invisible Unicode characters6 Sept
- Attackers Hijack MikroTik Routers Through Internet-Exposed SSH Without Authentication6 Sept
- Four REVSTEALER-Linked Modules Disable Windows Update and Defender to Run a Crypto Miner6 Sept
- Unpatched Magento and Adobe Commerce Zero-Day Exploited to Backdoor Online Stores5 Sept
- Attackers Breached JetBrains Cadence via Unpatched TeamCity, Extracting AWS Credentials5 Sept
- Critical VMware Workstation and Fusion Flaw Lets VM Admins Execute Host Code5 Sept
- Over 5,400 hacked sites serve ClickFix payloads stored on the blockchain5 Sept
- Trezor Says ShipMonk Breach Exposed 67,000 U.S. Customers' Data It Said Was Deleted5 Sept
- OpenAI admits it didn't disclose rogue AI wiki hijacking incident5 Sept
- Thousands of OpenAI Agents Quietly Turned an Abandoned Wiki Into Their Coordination Channel5 Sept
- Attackers Exploit PaperCut Flaws to Steal Credentials From Schools and Universities5 Sept
- Government Rails Site Hit Hours After CVE Patch4 Sept
- IDScan sued over alleged data breach affecting 153 million drivers4 Sept
- Using a VM to Contain an AI Agent4 Sept
- Companies Have 6 Months to Prepare for Automated Attacks4 Sept
- Critical Citrix NetScaler auth bypass now leveraged in attacks4 Sept
- PostgreSQL Fixes 12-Year-Old Logical Decoding Flaw Enabling Replication-Role Code Execution4 Sept
- Dirty Frag (CVE-2026-43284): the Linux kernel bug that turns read access into root4 Sept
- New Ted Backdoor Hides Inside Victims' Own HAProxy Builds to Intercept Web Traffic4 Sept
- 39 New Methods That Compromise Passkey Authentication4 Sept
- New CrowdStrike 'FalconFlank' zero-day grants SYSTEM privileges4 Sept
- G7 urges organizations to prepare for quantum cyber threats4 Sept
- Compliance teams have gone continuous, but their evidence-gathering hasn’t caught up4 Sept
- [NEU] [mittel] Grafana Enterprise: Mehrere Schwachstellen ermöglichen Erlangen von Benutzer- oder Administratorrechten4 Sept
- [NEU] [kritisch] vm2: Mehrere Schwachstellen ermöglichen Codeausführung4 Sept
- [NEU] [hoch] Microsoft Clouddienste: Mehrere Schwachstellen4 Sept
- [NEU] [mittel] Dell integrated Dell Remote Access Controller: Schwachstelle ermöglicht Codeausführung4 Sept
- [NEU] [hoch] Dell Secure Connect Gateway: Mehrere Schwachstellen4 Sept
- [NEU] [mittel] Ollama: Schwachstelle ermöglicht Offenlegung von Informationen4 Sept
- [NEU] [mittel] MongoDB: Mehrere Schwachstellen4 Sept
- [NEU] [hoch] util-linux: Schwachstelle ermöglicht Privilegieneskalation4 Sept
- [NEU] [hoch] MISP: Mehrere Schwachstellen4 Sept
- [NEU] [UNGEPATCHT] [mittel] bluez: Schwachstelle ermöglicht Codeausführung4 Sept
- [NEU] [hoch] SEPPmail Secure E-Mail Gateway: Mehrere Schwachstellen4 Sept
- [NEU] [hoch] Kibana: Mehrere Schwachstellen4 Sept
- Insurers Search for Answers to Rein in Rogue AI4 Sept
- Google warns of new Chrome zero-day flaw exploited in attacks4 Sept
- [UPDATE] [hoch] Intel Prozessoren: Mehrere Schwachstellen4 Sept
- [UPDATE] [hoch] Red Hat Enterprise Linux (libsoup): Mehrere Schwachstellen4 Sept
- [UPDATE] [hoch] GnuPG: Mehrere Schwachstellen4 Sept
- [UPDATE] [hoch] Net-SNMP: Schwachstelle ermöglicht Codeausführung und DoS4 Sept
- [UPDATE] [hoch] Linux Kernel: Mehrere Schwachstellen ermöglichen Denial of Service4 Sept
- [UPDATE] [mittel] Grub: Mehrere Schwachstellen4 Sept
- [UPDATE] [hoch] Linux Kernel: Mehrere Schwachstellen4 Sept
- [UPDATE] [hoch] Linux Kernel: Mehrere Schwachstellen4 Sept
- [UPDATE] [mittel] OpenSSL und LibreSSL: Mehrere Schwachstellen4 Sept
- [UPDATE] [hoch] Linux Kernel: Mehrere Schwachstellen ermöglichen Denial of Service4 Sept
- [UPDATE] [hoch] Linux Kernel: Mehrere Schwachstellen4 Sept
- [UPDATE] [hoch] WebKitGTK: Mehrere Schwachstellen4 Sept
- [UPDATE] [hoch] Linux Kernel: Mehrere Schwachstellen ermöglichen Denial of Service4 Sept
- [UPDATE] [mittel] Perl: Schwachstelle ermöglicht Codeausführung und Offenlegung von Informationen4 Sept
- [UPDATE] [mittel] OpenSSL: Schwachstelle ermöglicht Umgehen von Sicherheitsvorkehrungen4 Sept
- [UPDATE] [hoch] WebKitGTK: Mehrere Schwachstellen4 Sept
- [UPDATE] [hoch] Linux Kernel: Mehrere Schwachstellen4 Sept
- [UPDATE] [hoch] Linux Kernel: Mehrere Schwachstellen4 Sept
- [UPDATE] [hoch] Apache Tomcat und Tomcat Native: Mehrere Schwachstellen4 Sept
- [UPDATE] [mittel] OpenSSL: Schwachstelle ermöglicht Denial of Service und Remote-Code-Ausführung4 Sept
- [UPDATE] [mittel] systemd: Mehrere Schwachstellen4 Sept
- [UPDATE] [hoch] WebKitGTK: Mehrere Schwachstellen4 Sept
- Security Vulnerability in a Voting System4 Sept
- Protecting Against Zero-Click Attacks4 Sept
- AI Coding Agents Are Installing Unknown/Untrusted Code on Corporate Networks4 Sept
- Angry Birds: Toy Ghouls’ new toys4 Sept
- [UPDATE] [hoch] Red Hat Enterprise Linux (DBI, perl-GD): Mehrere Schwachstellen4 Sept
- [UPDATE] [mittel] docker: Schwachstelle ermöglicht Manipulation von Dateien4 Sept
- [UPDATE] [hoch] libssh2: Mehrere Schwachstellen4 Sept
- [UPDATE] [mittel] ffmpeg: Mehrere Schwachstellen ermöglichen Codeausführung und DoS4 Sept
- [UPDATE] [hoch] ffmpeg: Mehrere Schwachstellen4 Sept
- Version 1.0: Deutsche Institutionen über TerminalFix-Kampagne kompromittiert4 Sept
- GPT-6 Astra Scores 100% on ExploitBench as OpenAI Blocks PoC Exploit Requests4 Sept
- US offers $10 million for info on Iranian allegedly behind cyberattacks on critical infrastructure4 Sept
- Selfie-Capture Dynamics as an Auxiliary Signal Against Deepfakes and Injection Attacks for Mobile Identity Verification4 Sept
- Deep-Research Agents Can Be Poisoned via User-Generated Content4 Sept
- How Reliable Is the Multi-Input Heuristic for Bitcoin Address Clustering in Law Enforcement Contexts?4 Sept
- Sealing the Audit-Runtime Gap for LLM Skills4 Sept
- When Optimization Becomes Manipulation: Defending Generative Search against Malicious Generative Engine Optimization4 Sept
- Privacy Leakage in Federated Learning: Gradient-Based Client Identity Inference and Defenses for Inertial Sensing in Vehicular Edge Networks4 Sept
- Boundary-Mutation Testing for Pattern-Based Secret Detection: A Rule-Level Method and Cross-Scanner Evaluation4 Sept
- Population-Calibrated Graph Screening at 835-Million-Address Scale, with Label-Free Transfer to New Chains4 Sept
- Differentially private federated learning with Byzantine-robust aggregation: A cross-domain framework for secure model training in banking and healthcare systems4 Sept
- A Bayesian Correlated Equilibrium for Early Insider-Threat Detection4 Sept
- Trust Me, I'm Your Developer: Self-Issued Authentication in Large Language Models4 Sept
- After Cheap Discovery: From unknown to known-and-unfixed4 Sept
- Inferring Hidden User Models from the Behavior of Personalized LLM Agents4 Sept
- Beyond the Trust Boundary: A Critical Reassessment of the FIDO2 Threat Model4 Sept
- Rent-a-RAG: Embedding-Space Watermarks for Auditing Third-Party RAG4 Sept
- AlcaTRAz - Anchored Tree-Rule Defense Against Jailbreaks4 Sept
- The Native-Signature Boundary in Post-Quantum Distributed Authorization4 Sept
- CRAW: Codec Robust Audio Watermarking4 Sept
- Beyond Small Patches: Black-Box Detection and Purification of Diverse Backdoor Triggers4 Sept
- Seeing Less Is Not Seeing Safely: Privacy Leakage from Task-Scoped Robot Perception Exports4 Sept
- Long-Range Indirect Control-Flow Prediction in Stripped Binaries via Dual Virtual Hubs and Multi-Task Graph Learning4 Sept
- SENTINEL-RL: Offloading Topological Reasoning from LLM Agents in the Security Operations Center4 Sept
- CASCADE: A Component Ablation and Corpus Audit of a Layered Local Defense for MCP-Based Systems4 Sept
- PatchBench: Evaluating AI Agents for Vulnerability Patching4 Sept
- Recognition Without Mitigation: Ethical Frameworks in Autonomous Offensive-LLM Agent Research4 Sept
- Measuring Harmfulness of Computer-Using Agents4 Sept
- Computer Science Conferences Should Require Nonrepudiable Experimental Results4 Sept
- Identifying AI Web Scrapers Using Canary Tokens4 Sept
- A Blind Trust, the Bloody Thrust: When Attacker-Controlled Hook Updates Steer AI Agent Harnesses towards Malicious Behaviors4 Sept
- Security in the Age of AI Teammates: An Empirical Study of Agentic Pull Requests on GitHub4 Sept
- AI-Assisted Design of a Post-Quantum Cryptographic Accelerator: A Deployed-Silicon Case Study4 Sept
- A Black Box for Agentic Processes: Blockchain-Anchored Evidence for AI Agent Communication, Human Oversight, and GRC Audits4 Sept
- Practice Makes (Im)Perfect: A Look Back at Benchmarking Practices for Microarchitectural Side-Channel Attacks4 Sept
- Flip, Don't Shuffle: Watermarking LLMs at the Speed of Inference4 Sept
- Refusal Before Decoding: Detecting and Exploiting Refusal Signals in Intermediate LLM Activations4 Sept
- Safety Training Modulates Harmful Misalignment Under On-Policy RL, But Direction Depends on Environment Design4 Sept
- Beyond Reproducibility: Towards Security-Aware Evaluation of Research Artifacts4 Sept
- NACRE: Rethinking Confidential Containers through Native Architectural Support4 Sept
- Privacy, Robustness, and Fairness Trade-offs in Federated Intrusion Detection: Geometric Indistinguishability at the Aggregation Interface4 Sept
- Spruce: Scalable Private Outsourced Retrieval Using Compact Embeddings4 Sept
- SecDT: A Profile-Based Security Layer for TRDP Communications4 Sept
- French hospital fined €500,000 after breach exposes data of 727,0003 Sept
- Incident response guide for AWS CloudTrail investigations – Part 23 Sept
- Incident response guide for AWS CloudTrail investigations – Part 13 Sept
- Large Enterprises Targeted in Fake Merger & Acquisition Scams3 Sept
- Coder's registry infrastructure compromised to push malicious modules3 Sept
- Large group of Serbian opposition, activist figures targeted with spyware3 Sept
- HPE patches critical ArubaOS-CX remote code execution flaw3 Sept
- ThreatsDay: CEO Phishing Kits, 5K Dropbox Account Hacks, OAuth Traps + 17 More Stories3 Sept
- MECCHA CHAMELEON can't hide from the RCE3 Sept
- Critical Cisco Nexus 9000 Flaw Lets Unauthenticated Remote Attackers Run Code as Root3 Sept
- BraZetsu Malware Turns Compromised Windows Hosts Into Criminal Marketplace Inventory3 Sept
- Critical Elementor Pro flaw exploited to take over WordPress sites3 Sept
- Your Employee’s Password Appeared in an Infostealer Log. Now What?3 Sept
- US and Canadian court data exposed in Thomson Reuters breach3 Sept
- 'Breeze Comet' Tears Into Brazilian & Global Financial Systems3 Sept
- US Becomes Top Target in RMM Phishing Campaign Spanning 46 Countries3 Sept
- [NEU] [mittel] Sonatype Nexus Repository Manager: Mehrere Schwachstellen ermöglichen Denial of Service3 Sept
- [UPDATE] [mittel] Linux Kernel: Mehrere Schwachstellen3 Sept
- [NEU] [hoch] Hitachi Energy RTU500: Mehrere Schwachstellen3 Sept
- [NEU] [hoch] Jenkins: Mehrere Schwachstellen3 Sept
- [NEU] [hoch] n8n: Mehrere Schwachstellen3 Sept
- [NEU] [kritisch] Langflow: Schwachstelle ermöglicht Ausführen von beliebigem Programmcode mit Administratorrechten3 Sept
- [NEU] [hoch] util-linux: Mehrere Schwachstellen3 Sept
- [NEU] [UNGEPATCHT] [hoch] RPM: Schwachstelle ermöglicht Codeausführung3 Sept
- [NEU] [hoch] F5 BIG-IP: Mehrere Schwachstellen3 Sept
- [NEU] [UNGEPATCHT] [hoch] zlib: Schwachstelle ermöglicht Denial of Service und Integritätsverlust3 Sept
- [NEU] [mittel] Linux Kernel: Mehrere Schwachstellen3 Sept
- [NEU] [mittel] Cisco Secure Email Gateway: Mehrere Schwachstellen ermöglichen Umgehen von Sicherheitsvorkehrungen3 Sept
- Plex warns users to patch security vulnerabilities immediately3 Sept
- Attackers Turn Trusted Node.js Runtime Into Malware Delivery Tool in Targeted Attacks3 Sept
- Shai-Hulud's Reach Just Grew to 469 Credential Locations. Here's What That Means3 Sept
- Hijacked ScreenConnect Installs Are Spreading Malware Like a Worm, Huntress Warns3 Sept
- Attackers Expose Ongoing AI Tool Use Targeting Organizations in Latin America3 Sept
- [UPDATE] [mittel] CoreDNS: Mehrere Schwachstellen ermöglichen Denial of Service3 Sept
- [UPDATE] [hoch] NGINX und NGINX Plus: Mehrere Schwachstellen3 Sept
- Pegasus Zero-Click Spyware Exploit Infects Serbian Student Movement Member's iPhone3 Sept
- [UPDATE] [hoch] Mozilla Firefox und Thunderbird: Mehrere Schwachstellen3 Sept
- CISA Adds Seven Exploited Flaws as Attackers Deploy Reverse Shells and Crypto Miners3 Sept
- Removing Speech, Keeping Activities: A Privacy Firewall for Acoustic Sensing in Assisted Living3 Sept
- Counter-GEO-Bench: Evaluating Defenses Against Information-Distorting Generative Engine Optimization3 Sept
- Transfer Safety Awareness for Cross-Modal Safety Drift in Multimodal Large Language Models3 Sept
- ClaimReceipt: Verifying Evidence Sufficiency and Coverage in Agent Evaluations3 Sept
- Modelstamp: Pre-Deserialization Verification of Machine-Learning Artifacts and Runtime Environment State3 Sept
- Barriers to Using Static Application Security Testing (SAST) Tools: A Literature Review3 Sept
- Overcoming the Randomness-Utility Trade-off in Answering Differentially Private Linear Queries3 Sept
- When Does Authorization End? Effect Closure at Provider Boundaries3 Sept
- CodePoisonRAG: Knowledge Poisoning Attacks on Retrieval-Augmented Code Generation3 Sept
- Agent Memory Is a Surface for Endogenous Authorization Laundering3 Sept
- SPADE: SPaT Attack Detection from the Connected Vehicle's Perspective3 Sept
- Agent Tools Orchestration Leaks More: Dataset, Benchmark, and Mitigation3 Sept
- TraceGuard: Process-Guided Firewall against Reasoning Backdoors in Large Language Models3 Sept
- ACLE-MCP: Attested Capability Leases for Execution-Time Trust in Remote LLM Tool Use3 Sept
- HEAT: Faster Fully Homomorphic Inference via Approximations-Weights Co-Adaptation3 Sept
- Hearing the Whispers: Black-Box Membership Inference Attacks on Finetuned TTS Models3 Sept
- Public-Sharing Labels and Verbatim Field Egress in an MCP-to-A2A Agent Configuration: A Controlled Multi-Model Study3 Sept
- Skill-as-API: Confidential Multi-Agent Coordination for Agentic Software Engineering3 Sept
- Ranked by the Matcher: A Reproducibility Audit of Knowledge Graph Extraction from Threat Reports3 Sept
- Backdoor Attacks on Speech Emotion Recognition via TTS-Generated Poisoning3 Sept
- Quantum Blind Rotation for Fast Functional Bootstrapping3 Sept
- LLM Watermarking as Big Data Provenance: A Deployment-Oriented Systematization3 Sept
- Automated Vulnerability Injection in Smart Contracts Using Large Language Models3 Sept
- A Finger on the Scale: Covert Policy Steering through Agentic Skills3 Sept
- The Shape of Ownership: Verifying LLM Provenance through Semantic Structures3 Sept
- SafeEvolve: Harness-Policy Co-Evolution from Agent Experience for Safety Alignment3 Sept
- The Implications of Linguistic Illegibility for LLM Security3 Sept
- Stored Is Not Supported: Typed Provenance and Assertion Guardrails for Persistent AI Agents3 Sept
- Implicit Manipulation for Skill Selection in LLM Agents with Semantic Matching3 Sept
- Type-Directed, Secure-by-Construction Enclave Partitioning for LLVM3 Sept
- CAPTCHAs in the Agentic Era: Solvers That Learn from Every Encounter3 Sept
- Context Inference Attacks Without Jailbreaks3 Sept
- PrimSynth: An Agentic Approach to Discover, Validate, and Synthesize Exploit Primitives for Linux Kernel Vulnerabilities3 Sept
- Evaluating ML-based Intrusion Detection Systems: The Illusion of Model Efficacy3 Sept
- SpiderSapien: Client-Centric Web Crawler and Security Scanner3 Sept
- Can Risk-Based Alerting Mitigate Cybersecurity Alert Fatigue?3 Sept
- WeaveMark: Robust and Scalable Multi-bit LLM Watermarking via Coded Payload Spreading3 Sept
- Agent Flight Recorder: Tamper-Evident Audit Trails with On-Chain Anchoring for Long-Horizon Tool-Using Agents3 Sept
- Towards Behavior Tree-Guided Vulnerability Detection with Lightweight LLMs3 Sept
- Adversarial Vulnerabilities of Neural Biomarker Identification Systems3 Sept
- H1 2026 Malware Vulnerability Trends3 Sept
- Hackers exploit Sangoma Switchvox flaw to deploy reverse shells2 Sept
- AI Gives Cybercriminals a Dangerous Time Advantage2 Sept
- WordPress backup plugin flaw exposes millions of sites to takeover attacks2 Sept
- Google, Anthropic, and OpenAI Unveil Cyber AI Models, Safeguards, and Access Programs2 Sept
- Demystifying Agent Tradecraft: Introducing SpecterOps Skills2 Sept
- Health data of more than 9.5 million people leaked from Aesto record system2 Sept
- Hackers exploit critical JFrog Artifactory flaw to forge admin tokens2 Sept
- New pro-Ukraine hacker group targets Russian companies with custom ransomware2 Sept
- Malicious Apache Modules Hijack Brazilian Government Site Traffic to Push Betting Pages2 Sept
- Six curl CVEs after OpenAI and Anthropic came back with zero2 Sept
- Version 1.0: SonicWall SMA1000 - Schwachstellen werden aktiv ausgenutzt2 Sept
- Dropbox accounts breached through Lenovo email verification flaw2 Sept
- Sality, one of the longest-running botnets, finally gets disrupted2 Sept
- [NEU] [hoch] Microsoft GitHub Enterprise Server: Mehrere Schwachstellen2 Sept
- [NEU] [hoch] Coolify: Schwachstelle ermöglicht Codeausführung2 Sept
- [NEU] [hoch] Jolokia: Schwachstelle ermöglicht Umgehen von Sicherheitsvorkehrungen2 Sept
- [NEU] [mittel] FasterXML Jackson: Schwachstelle ermöglicht nicht spezifizierten Angriff2 Sept
- [NEU] [mittel] Erlang/OTP: Mehrere Schwachstellen2 Sept
- [NEU] [hoch] Kibana: Mehrere Schwachstellen2 Sept
- [NEU] [hoch] Rockwell Automation FactoryTalk Activation Manager und Historian Machine Edition: Mehrere Schwachstellen2 Sept
- [NEU] [mittel] Rockwell Automation ControlLogix 5580, CompactLogix 5380, und CompactLogix 5480: Mehrere Schwachstellen ermöglichen Denial of Service2 Sept
- [NEU] [hoch] Aruba ArubaOS-CX: Mehrere Schwachstellen2 Sept
- [NEU] [mittel] IBM QRadar SIEM: Schwachstelle ermöglicht Offenlegung von Informationen2 Sept
- [NEU] [mittel] IBM Java SDK: Schwachstelle ermöglicht Denial of Service2 Sept
- [NEU] [mittel] Filebeat: Schwachstelle ermöglicht Denial of Service2 Sept
- Microsoft Defender flags legitimate Google search links as malicious2 Sept
- [NEU] [hoch] Mozilla Firefox und Thunderbird: Mehrere Schwachstellen2 Sept
- [NEU] [hoch] Elasticsearch: Mehrere Schwachstellen2 Sept
- [UPDATE] [hoch] Red Hat Enterprise Linux (librest, pipewire): Mehrere Schwachstellen2 Sept
- [UPDATE] [hoch] Red Hat Enterprise Linux (Pillow): Mehrere Schwachstellen2 Sept
- [UPDATE] [mittel] Unbound: Mehrere Schwachstellen2 Sept
- [UPDATE] [mittel] Netty: Mehrere Schwachstellen2 Sept
- [UPDATE] [hoch] Ubuntu Linux (ubuntu-pro-client): Mehrere Schwachstellen2 Sept
- [UPDATE] [hoch] IBM License Metric Tool: Mehrere Schwachstellen ermöglichen nicht spezifizierten Angriff2 Sept
- [UPDATE] [mittel] FreeRDP: Mehrere Schwachstellen2 Sept
- [UPDATE] [hoch] FreeRDP: Mehrere Schwachstellen2 Sept
- [UPDATE] [mittel] Red Hat Enterprise Linux: Schwachstelle ermöglicht Privilegieneskalation und Offenlegung von Informationen2 Sept
- [UPDATE] [hoch] FreeRDP: Mehrere Schwachstellen2 Sept
- [UPDATE] [mittel] Red Hat OpenShift und OpenShift AI (urllib3): Schwachstelle ermöglicht Denial of Service2 Sept
- [UPDATE] [hoch] FreeRDP: Mehrere Schwachstellen2 Sept
- [UPDATE] [niedrig] OpenSSH: Mehrere Schwachstellen ermöglichen Codeausführung2 Sept
- [UPDATE] [mittel] libxml2 (exsltDynMapFunction): Schwachstelle ermöglicht Denial of Service2 Sept
- [UPDATE] [mittel] http/2 Implementierungen: Schwachstelle ermöglicht Denial of Service2 Sept
- [UPDATE] [hoch] Google Chrome und Microsoft Edge: Mehrere Schwachstellen2 Sept
- [UPDATE] [hoch] Google Chrome / Microsoft Edge: Mehrere Schwachstellen2 Sept
- [UPDATE] [hoch] Google Chrome / Microsoft Edge: Schwachstelle ermöglicht nicht spezifizierten Angriff2 Sept
- [UPDATE] [hoch] GNU libc: Schwachstelle ermöglicht Codeausführung2 Sept
- An AI-Assisted Cyber Attack: Inside a Unit 42 Investigation2 Sept
- US charges Russian for infecting 80,000 freelancers with malware2 Sept
- SonicWall warns of actively exploited SMA1000 zero-day flaws2 Sept
- A SoK for SoCs: Reading the TI Leaves on AI for Cyber Threat Intelligence Generation and Sharing2 Sept
- A Hybrid Insider Threat Detection Framework Combining Multi-Agent Simulation, Layered SIEM Correlation, and Theory-of-Mind Reasoning2 Sept
- Effective Interventions Against AI-Enhanced Scams2 Sept
- Using LLMs to Elicit Security Requirements for Service-Oriented Cyber Ranges2 Sept
- Influence of Logging Frameworks on Bind92 Sept
- AKRASIA: Stealthy Backdoor Attack on Reasoning-based Code LLMs2 Sept
- Long-Horizon State Tracking in LLMs: Executing MD5 through a Deep Sequence of Dependent Tool Calls2 Sept
- OpenAgentFlow: Enabling System-Wide Safety Boundaries for Heterogeneous AI Agent Fleets2 Sept
- HiveTraceGuard-Pro: A Compact Generative Guardrail for Prompt Injection, Jailbreaks, and Adversarial Obfuscation2 Sept
- JENGA: Exploiting Counter-Based RowHammer Countermeasures to Break Real-Time Predictability2 Sept
- Security Testing Framework for Web Applications: Benchmarking ZAP V2.12.0 and V2.13.0 by OWASP as an example2 Sept
- Reveree: Diagnosing LLM Reverse-Engineering Agents2 Sept
- What's in Your Agent's Context? Context Privilege Escalation Attacks against AI Agent Harness2 Sept
- Hidden Services Protocol for Mixnets2 Sept
- TRIS: A Tri-Layer Retrieval Integrity Sieve Against Knowledge Poisoning2 Sept
- EvoFlint: An Evolutionary Atlas of Multi-Turn LLM Vulnerabilities2 Sept
- Same Request, Different Boundary: Evaluating Cybersecurity Assistance across Conversational Contexts2 Sept
- Membership Inference in Fine-tuned Diffusion Language Models via Token-level Memorization Asymmetry2 Sept
- Lacan: Making Accountability in Anonymous Networks Real2 Sept
- Griotte: Verified Compartmentalisation via Capabilities2 Sept
- Sentinel-Based Failover for QKD-Augmented IPsec Tunnels2 Sept
- Athena: Vulnerability-Affected Library Identification via Knowledge Graph Completion2 Sept
- Does Runtime Topology Context Improve LLM-Generated Kubernetes Security Patches?2 Sept
- FedReview: Review and Dispose Poisoned Updates without Validation Datasets or Historic Knowledge2 Sept
- The Popularity Hypothesis in Software Security: A Large-Scale Replication with PHP Packages2 Sept
- GuidedBench: Measuring and Mitigating the Evaluation Discrepancies of In-the-wild LLM Jailbreak Methods2 Sept
- AgentProv: Auditing Agentic LLM API Providers via Tool-use Policy Probes2 Sept
- A Formal Analysis of Agent Payment Protocols2 Sept
- DUPIN: Attack Learning Is Still Needed! Demonstrating Few-Shot after Unsupervised Pretraining Is A Nimble Forensics Learner2 Sept
- Delegation Without Trust: An Empirical Gap Analysis of Identity, Authorization, and Runtime Governance in Multi-Agent LLM Systems2 Sept
- Workload Identification with Physical Side Channels for AI Governance2 Sept
- OreProof: Verifiable Provenance with Limited Disclosure for Critical-Minerals Supply Chains Using Zero-Knowledge Proofs2 Sept
- Identification of Compositional Risks in Data Protection Impact Assessments and Beyond2 Sept
- NeuroPriv: Adversarial Representation Learning for Privacy in Wearable EEG Systems2 Sept
- Federated Trust for Embodied Robot Capability Marketplaces2 Sept
- Don't Trust the Code, Check Its Effects: Runtime Refinement for Regenerated Systems Code Under an Adversarial Generator2 Sept
- Capability-Gated Language Models: Security Composes, Utility Does Not2 Sept
- Does Reasoning Mitigate Backdoor Attacks? A Neuro-Symbolic Perspective2 Sept
- SoK: When Safe Agents Fail Together: The Security of Multi Agent LLM Systems2 Sept
- NeuroGraph: An AI Graph-Driven Neuro-Symbolic Framework for Explainable Threat Reasoning in Advanced Manufacturing2 Sept
- Automating Static Code Analysis Through CI/CD Pipeline Integration2 Sept
- PhantomCall: Evading ML Malware Detectors via Function Call Graph Perturbation2 Sept
- GlitchLab: A Hardware-in-the-Loop Optimizer for Physical Fault Injection2 Sept
- The Safeguard Worked. Is the LLM System Safer?2 Sept
- Transferable End-to-End Optimization for Indirect Long-Term Memory Poisoning in LLM Agents2 Sept
- When Safety Routing Breaks: Understanding Alignment Fragility under Benign Fine-Tuning2 Sept
- RISA: Response Inspection and Selective Actions for Refusal Calibration in Large Language Models2 Sept
- One Prompt Is Enough: Watermark Laundering Through Foundation Image Models2 Sept
- Defense-as-Skill: Evolving Runtime Guard Skill for Skill-Augmented Agents2 Sept
- MultiGait: A Multi-Sensor Multi-Perspective Multi-Session Biometric Inference Benchmark and its Dataset2 Sept
- Old, Unpatched Flaws Give Attackers Access to Philippines Nuclear Agency2 Sept
- Improving our alignment and security efforts1 Sept
- Hackers abuse Faronics Deploy admin tool to install ScreenConnect1 Sept
- AI Model Evaluator METR Hit by Credential Theft, Probing1 Sept
- Critical Langflow flaw exploited to steal OpenAI and AWS keys1 Sept
- Forescout Research Tests Whether AI Can Create PLC Attacks1 Sept
- Leaked Russian Cyber-Operations Training Materials1 Sept
- Hackers push malicious Virtualizor update in BGP hijacking attack1 Sept
- The dark figure of supply chain detection1 Sept
- 13 Malicious Packagist Packages Target Unpatched iPhones to Steal Crypto Wallet Seeds1 Sept
- ClickFix Campaign Compromises 31 Orgs, Abuses Polygon Blockchain1 Sept
- Iranian Hackers Pose as Recruiters to Deliver Cross-Platform RATs Through Coding Tests1 Sept
- Filigran Adds AI-Powered Attack Chaining to OpenAEV for Autonomous Pentesting1 Sept
- Nearly 22,000 Microsoft Exchange servers vulnerable to hijack attacks1 Sept
- Iranian cyber spies target aviation, fintech developers with new malware1 Sept
- Cyber risk from frontier AI poses ‘most immediate concern’ to global financial system, watchdog warns1 Sept
- Threat Actors Don’t Want Better Attacks. They Want Repeatable Ones1 Sept
- [UPDATE] [hoch] xz: Schwachstelle ermöglicht Codeausführung1 Sept
- [UPDATE] [hoch] docker: Mehrere Schwachstellen1 Sept
- [UPDATE] [mittel] VMware Tanzu Spring Framework (MVC and WebFlux): Mehrere Schwachstellen1 Sept
- [UPDATE] [hoch] VMware Tanzu Spring Security: Schwachstelle ermöglicht Umgehen von Sicherheitsvorkehrungen1 Sept
- [UPDATE] [mittel] systemd: Mehrere Schwachstellen1 Sept
- [UPDATE] [hoch] Atlassian Bamboo, Bitbucket, Confluence, Fisheye, Crucible, Jira und Jira Service Management: Mehrere Schwachstellen1 Sept
- [UPDATE] [hoch] Red Hat Ansible Automation Platform (node-tar, linkify-it, protobufjs, brace-expansion, fast-uri, DOMPurify): Mehrere Schwachstellen1 Sept
- [UPDATE] [hoch] libssh: Mehrere Schwachstellen1 Sept
- [UPDATE] [mittel] Red Hat Enterprise Linux (python-wheel): Schwachstelle ermöglicht Privilegieneskalation und Codeausführung1 Sept
- [UPDATE] [mittel] expat: Schwachstelle ermöglicht Codeausführung1 Sept
- [UPDATE] [mittel] FasterXML Jackson: Mehrere Schwachstellen1 Sept
- [UPDATE] [mittel] Keycloak (netty-codec-http): Schwachstelle ermöglicht Manipulation von Dateien1 Sept
- [UPDATE] [hoch] Red Hat Enterprise Linux (urllib3): Mehrere Schwachstellen ermöglichen Denial of Service1 Sept
- [UPDATE] [hoch] Shibboleth Service Provider: Schwachstelle ermöglicht SQL Injection1 Sept
- [UPDATE] [mittel] VMware Tanzu Spring Framework und Spring Security: Mehrere Schwachstellen ermöglichen Umgehen von Sicherheitsvorkehrungen1 Sept
- [UPDATE] [hoch] libxml2: Mehrere Schwachstellen ermöglichen Denial of Service1 Sept
- [UPDATE] [hoch] Angular: Mehrere Schwachstellen1 Sept
- [UPDATE] [hoch] Linux Kernel: Mehrere Schwachstellen1 Sept
- [UPDATE] [hoch] VMware Tanzu Spring Security: Schwachstelle ermöglicht Umgehen von Sicherheitsvorkehrungen1 Sept
- [UPDATE] [hoch] Google Chrome und Microsoft Edge: Mehrere Schwachstellen1 Sept
- [UPDATE] [hoch] Google Chrome/Microsoft Edge: Mehrere Schwachstellen1 Sept
- [UPDATE] [hoch] Linux Kernel: Mehrere Schwachstellen1 Sept
- [UPDATE] [hoch] vllm: Schwachstelle ermöglicht Umgehen von Sicherheitsvorkehrungen1 Sept
- [UPDATE] [hoch] Atlassian Bamboo, Bitbucket, Confluence, Fisheye, Crucible, Jira und Jira Service Management: Mehrere Schwachstellen1 Sept
- [UPDATE] [hoch] Angular: Mehrere Schwachstellen1 Sept
- [UPDATE] [hoch] Google Chrome / Microsoft Edge: Mehrere Schwachstellen1 Sept
- [UPDATE] [kritisch] Kemp LoadMaster: Mehrere Schwachstellen1 Sept
- [UPDATE] [hoch] WebKitGTK: Mehrere Schwachstellen1 Sept
- [UPDATE] [hoch] Samba: Mehrere Schwachstellen1 Sept
- [UPDATE] [hoch] Samba: Mehrere Schwachstellen1 Sept
- [UPDATE] [mittel] Elasticsearch und Kibana: Mehrere Schwachstellen1 Sept
- [UPDATE] [hoch] Red Hat Enterprise Linux: Schwachstelle ermöglicht Umgehen von Sicherheitsvorkehrungen1 Sept
- [UPDATE] [mittel] GnuTLS: Mehrere Schwachstellen1 Sept
- [UPDATE] [hoch] Linux Kernel (Fragnesia): Schwachstelle ermöglicht Erlangen von Administratorrechten1 Sept
- [UPDATE] [mittel] GnuTLS: Mehrere Schwachstellen ermöglichen Umgehen von Sicherheitsvorkehrungen1 Sept
- [UPDATE] [mittel] Samba: Mehrere Schwachstellen1 Sept
- [UPDATE] [mittel] libtasn1: Schwachstelle ermöglicht Denial of Service1 Sept
- [UPDATE] [mittel] GnuTLS: Schwachstelle ermöglicht Denial of Service1 Sept
- Russia-Aligned UAC-0099 Plants Nuclear Weapon Prompt in Malware to Disrupt AI Analysis1 Sept
- Recently patched PaperCut zero-days used in data theft attacks1 Sept
- Curvature Cryptanalysis of Smooth Transformer Feed-Forward Networks1 Sept
- Ouroboros: Self-Referential Backdoor Attacks on Speech Enhancement via Clean Audio Triggers1 Sept
- Building the Truman Show: A TrustZone-Based Framework for Lightweight Out-of-band Kernel Security Monitoring1 Sept
- JITterFlip: Uncovering Fault Attack Surfaces in JIT-Compiled LLM Serving1 Sept
- Reactive Peripheral Modeling for Faithful Firmware Rehosting1 Sept
- OASIS: Optimizing Attacker Sequences for Hard-Label Black-Box Text Attacks1 Sept
- Learning diverse attacks on large language models for robust red-teaming and safety tuning1 Sept
- Benchmark Contamination: A Taxonomy Organized by Defeated Mitigation1 Sept
- Safe to Resume? Breaking Execution Continuity of Agent Execution via Rollback1 Sept
- A Multi-Month Study of Git Commit Signing1 Sept
- WoE Wrote It? Watermarking Mixture-of-Experts LLMs for Black-Box Text Provenance1 Sept
- Not the Same Protector: Deployment-Dependent Protective Intervention in LLMs1 Sept
- Mechanizing Typed Regulatory Actions for Security Tokens: Semantics, Falsification, and Bounded EVM Evidence1 Sept
- Auditing and Mitigating Privacy Leakage in Cloud-Edge Collaborative Decoding1 Sept
- Temporal Analysis of NetFlow Datasets for Network Intrusion Detection Systems1 Sept
- Watch your steps: Dormant Adversarial Behaviors that Activate upon LLM Finetuning1 Sept
- Relevance as a Vulnerability: How Web Retrieval Degrades Safety Alignment in LLM Agents1 Sept
- Hidden Thoughts Are Not Secret: Reasoning Trace Exposure in LLMs1 Sept
- Enhancing Web Application Firewalls with BERT-GNN for SQL Injection Detection1 Sept
- Identity by Design, Demographics by Accident: Demographic Leakage and Suppression in Behavioral Biometric Embeddings1 Sept
- Membership is Ownership: A Robust Ownership Verification Framework for Diffusion Models1 Sept
- CARVY-FL: Client Anticlustering for Robust Voting in Provably Secure Federated Learning1 Sept
- Zero-Knowledge Predicate Proofs Between AI Agents: A Measured, Cross-Protocol Gateway and the Source-Integrity Gap1 Sept
- Reachability-Based Capability Confinement for LLM Agents under Indirect Prompt Injection1 Sept
- ActReal: System-Level Mobile Agents Challenge Mobile Automation Detection1 Sept
- A Comprehensive Survey on Linguistic Steganography: Methods, Countermeasures, Evaluation, and Challenges1 Sept
- Dataset Protection via Watermarked Canaries in Retrieval-Augmented LLMs1 Sept
- Auditing Anonymous AI Models: A Four-Stage Protocol for Black-Box Identity Verification1 Sept
- Authority-Inference Separation in Agentic Finance: First-Line Control, Blockchain Enforcement, and Replayable Assurance1 Sept
- Beyond Object Authentication: Context-Closed Post-Quantum Authentication for the WebPKI1 Sept
- Breaking Ambient Trust: In-Network Per-Process Access Control Against Lateral Movement1 Sept
- A New Algebraic Algorithm for LWE1 Sept
- Influence Is Not Authority: When Causal Guardrail Signals Make Legitimate Tool Use Look Like an Attack in Tool-Using LLM Agents1 Sept
- GhostSplat: Input-Triggered Backdoors for Multi-View-Consistent 3D Content Manipulation in Feed-Forward Gaussian Splatting1 Sept
- SIR: Self-improving Red-teaming for Compute Use Agents1 Sept
- The Exclusion Ratchet: False-Positive Suppression Accumulates and Persists in Detection Rule Repositories1 Sept
- Understanding Stage-Wise Utility-Risk Trade-offs in LLM Agent Memory1 Sept
- Reducio: Optimized Confidential Serverless Cloud Deployments for Enterprise Customers1 Sept
- Balancing Privacy, Utility, and Safety in LLM Alignment through Preference Optimization1 Sept
- Drishti: AI-Led Human-Directed Vulnerability Auditing for 5G Cores1 Sept
- Extracting Knowledge from Tools in LLM Agents1 Sept
- A Roadmap to Available ICS Datasets and Testbeds for Cybersecurity Research1 Sept
- POLYFLOW: A Neuro-Symbolic Framework for Static Cross-Language Information Flow Analysis1 Sept
- Using Hyper-V Sockets for Real-time Data Extraction from a Malware Analysis Sandbox1 Sept
- Attesting Outputs and Delegation Ancestry in Multi-Agent AI Systems1 Sept
- Why Are LLM Backdoor Defenses Fragmented? A Feature-Level Explanation with Sparse Autoencoders1 Sept
- Disappearing Ink: Obfuscation Breaks N-gram Code Watermarks in Theory and Practice1 Sept
- SoK: Systematizing Generation, Characteristics, and Defenses in LLM-Generated Phishing1 Sept
- SUB-PLAY: Adversarial Policies against Partially Observed Multi-Agent Reinforcement Learning Systems1 Sept
- MTD-Playground: An Attacker-Aware Evaluation Framework for Network Moving Target Defense1 Sept
- Breaking the Code: Security Assessment of AI Code Agents Through Systematic Jailbreaking Attacks1 Sept
- HSMLog: Small Language Model-Assisted Hardware Security Module Log Anomaly Detection with Behavioral Analysis1 Sept
- Privacy-Preserving LLM Embedding Transmission for End-Cloud Collaboration1 Sept
- The Value of Spike Timing: A Leakage-Resistant Benchmark of SNN Design Choices for Network Intrusion Detection1 Sept
- Breaking Darknet CAPTCHAs with general purpose LLM1 Sept
- Privacy-Enhanced Zero-Order Federated Learning via xMK-CKKS over Wireless Channels1 Sept
- Trust Under Siege: Label Spoofing Attacks against Machine Learning for Android Malware Detection1 Sept
- A Comprehensive Study of Native Code Bugs in Python Applications1 Sept
- A Simple Transformer Pipeline for Full-Key Side-Channel Attacks on Uncropped Datasets1 Sept
- Adversarial Trust Poisoning in Vehicular Collaborative Perception1 Sept
- ARMOR: Manifold-Oriented Training for Adversarially Robust Aerial Object Detection under Data Scarcity1 Sept
- Secret Stealing Attacks on Local LLM Fine-Tuning through Supply-Chain Model Code Backdoors1 Sept
- Can escalation channels redirect reward hacking toward defect disclosure?1 Sept
- Towards Operator-Empowered Vulnerability Hotfixing for 5G Radio Access Networks1 Sept
- ECLIPSE: Self-Evolving Stealthy Prompt Injection Attack against Long-Horizon Agentic Systems1 Sept
- Lie to Me: Finding Bugs in ZK DSL Toolchains with Adversarial Witness Injection1 Sept
- Resolving Availability and Run-time Integrity Conflicts in Real-Time Embedded Systems1 Sept
- BadPatches: Routing-Aware Backdoor Attacks on Vision Mixture-of-Experts1 Sept
- Beyond the Payload: How User Invocation Shapes Coding Agent Vulnerability to Repository Poisoning1 Sept
- SingProbe Technical Report1 Sept
- MIRAGE: Misleading Retrieval-Augmented Generation via Black-box and Query-agnostic Poisoning Attacks1 Sept
- AgenTRIM: Tool Risk Mitigation for Agentic AI1 Sept
- More Haste, Less Speed: Weaker Single-Layer Watermark Improves Distortion-Free Watermark Ensembles1 Sept
- Breaking MCP with Function Hijacking Attacks: Novel Threats for Function Calling and Agentic Models1 Sept
- Unlearning on Spatio-Temporal Graphs through Subgraph Virtual Edge Reconstruction1 Sept
- The Fragility of Jailbreak Robustness Across Operational States1 Sept
- "Setting up TLS authentication was hell": A Usability Study of Client Certificate Authentication1 Sept
- The Web-CLI: Verifiable Privacy for Tools, Models, and Inference Engines in the Browser1 Sept
- VeriX-Anon: A Multi-Layered Framework for Mathematically Verifiable Outsourced Target-Driven Data Anonymization1 Sept
- DP-VOXLET: Provable Speaker Anonymization for Disentangled Speech Representations1 Sept
- T-MAP: Red-Teaming LLM Agents with Trajectory-aware Evolutionary Search1 Sept
- The Agentic SOC – From AI Theater to Real Defense1 Sept
- Is Someone Hacking DoD Refrigerators?31 Aug
- North Korean Job Fraud Expands Beyond IT Into Healthcare and Sales31 Aug
- Microsoft Exchange Online outage causes email failures, auth issues31 Aug
- Chinese Fire Ant hackers turn Cisco routers into spying platforms31 Aug
- Berlin confirms data theft after Rhysida ransomware attack claims31 Aug
- Pharmaceutical giant McKesson warns of 'service degradation' following cyberattack31 Aug
- Aurora Ransomware Operators Use Cursor AI in Attacks Against 10 Targets31 Aug
- [NEU] [UNGEPATCHT] [mittel] Red Hat Undertow: Schwachstelle ermöglicht Denial of Service31 Aug
- Securing Claude Code: The New Compliance API, Local Visibility, and Identity Governance31 Aug
- [NEU] [mittel] IBM DataPower Gateway: Mehrere Schwachstellen31 Aug
- [NEU] [hoch] Langflow OSS: Mehrere Schwachstellen31 Aug
- [NEU] [mittel] sudo: Schwachstelle ermöglicht Umgehen von Sicherheitsvorkehrungen31 Aug
- [NEU] [hoch] Apache Wicket: Mehrere Schwachstellen31 Aug
- Hiding Prompt Injection in Legal Filing31 Aug
- [UPDATE] [mittel] Keycloak: Mehrere Schwachstellen ermöglichen Umgehen von Sicherheitsvorkehrungen31 Aug
- [UPDATE] [mittel] Keycloak: Mehrere Schwachstellen ermöglichen Offenlegung von Informationen31 Aug
- [UPDATE] [mittel] Keycloak: Mehrere Schwachstellen31 Aug
- [NEU] [mittel] MariaDB Connectors: Mehrere Schwachstellen31 Aug
- [NEU] [mittel] Graylog: Mehrere Schwachstellen31 Aug
- [NEU] [mittel] Flowise: Mehrere Schwachstellen ermöglichen Umgehen von Sicherheitsvorkehrungen und DoS31 Aug
- [NEU] [hoch] Microsoft Edge: Mehrere Schwachstellen31 Aug
- Spring Ring: An Inside Look at Voice Phishing Campaigns in Microsoft Teams31 Aug
- ValleyRAT masquerading as adware31 Aug
- [NEU] [hoch] MongoDB BI Connector und ODBC Driver: Mehrere Schwachstellen31 Aug
- [NEU] [mittel] vllm: Mehrere Schwachstellen ermöglichen Denial of Service31 Aug
- Microsoft asks users to ignore 'Antivirus is turned off' errors31 Aug
- [UPDATE] [mittel] Linux Kernel: Mehrere Schwachstellen31 Aug
- [UPDATE] [hoch] QEMU: Schwachstelle ermöglicht Privilegieneskalation31 Aug
- [UPDATE] [mittel] Roundcube: Mehrere Schwachstellen31 Aug
- [UPDATE] [hoch] Linux Kernel: Mehrere Schwachstellen31 Aug
- [UPDATE] [hoch] Linux Kernel: Mehrere Schwachstellen31 Aug
- [UPDATE] [mittel] Linux Kernel: Schwachstelle ermöglicht Privilegieneskalation31 Aug
- SkillSafetyBench: Evaluating Agent Safety under Skill-Facing Attack Surfaces31 Aug
- CAITLYN: Can LLM Agents Autonomously Synthesize Defenses against Emerging Injection Attacks?31 Aug
- A Wolf in Sheep's Clothing: Targeted Routing Hijacking in Federated RAG31 Aug
- Securing Multi-Agent GIS Systems: Risk Evaluation and Prompt Hardening Optimization31 Aug
- Balancing the privacy-utility trade-off: How to draw reliable conclusions from private data31 Aug
- FlyBlind: Cross-Slice Timeliness Attacks on UAV Situational Awareness over 5G31 Aug
- Semantic Watermarking with Order-Robust Detection over Sub-sentence Units31 Aug
- DisCTI: Who Needs to Know Timely? Automated Sector-Aware Cyber Threat Intelligence Dissemination31 Aug
- Moirae: A Multimodal Agent Collaborative Framework for Dynamic Android Malware Detection31 Aug
- Can Tainted Pixels Expose Deepfake Videos?31 Aug
- Not to Break, but to Attest: Adversarial Probes for Privacy-Preserving LLM Verification31 Aug
- A User-Centric Context-Aware Permission Governance Framework for Privacy Control in Default Mobile Applications31 Aug
- Revisiting Continuous Noise Sampling for Multi-Party Differential Privacy31 Aug
- Fully Unleashing the Multimodal Attacker: Meta-Adaptive Jailbreaking of Vision-Language Models31 Aug
- FISGuard: Defending Against Membership Inference via Fixed Input Subspaces31 Aug
- ANCHOR: A Vision for Secure Persistent Key-Value Stores in Disaggregated Data Centers31 Aug
- Circuit Discovery Helps Detect LLM Jailbreaking: A Mechanistic Interpretability Study31 Aug
- The Impact of Magma: A Ground-Truth Fuzzing Benchmark31 Aug
- Memorization Is Not Extraction: Tight Differential-Privacy Bounds and Audit Blind Spots31 Aug
- Exploiting Per-Core Leakage: Electromagnetic Side-Channel Monitoring of Multicore Architectures31 Aug
- Compared to What? A Human-Anchored Security Benchmark for LLM-Generated Infrastructure-as-Code31 Aug
- LongPIBench: A Long-Context Benchmark for Prompt Injection31 Aug
- ContextLeak: Exfiltrating LLM Agent Context via Malicious Tools31 Aug
- CamoDocs: A Poisoning Attack Against Retrieval-Augmented Language Models Using Camouflaged Documents31 Aug
- BEACON: Behavior-Anchored Cross-Source Knowledge Graph Construction for Cyber Threat Intelligence31 Aug
- TagZilla: Automated Owner and Abuse Type Tagging for Indicators of Compromise in Threat Reports31 Aug
- Layered LLM Defenses as an Ensemble: Access Tiers, Inference Cost, and the Measured Failure Correlation Between Defense Layers31 Aug
- When Verified Source Becomes Attack Input: Defending Smart Contracts Against LLM-Based Vulnerability Scanning31 Aug
- Quantum-Based Solutions for Security Enhancement in Open Radio Access Networks31 Aug
- Recognition Without Enforcement: Configuration-Dependent Failures in LLM Agent Instruction Arbitration and External Control31 Aug
- Quantization-Triggered Backdoors in Language Models: Cross-Quantizer Transferability and the Validation--Deployment Gap31 Aug
- REPLICANT: Learning Policies for Evading and Hardening Malware Detectors31 Aug
- CHISEL-ing Back Source Code with AI-enabled Iterative Recovery31 Aug
- GSPR: Aligning LLM Safeguards as Generalizable Safety Policy Reasoners31 Aug
- GREAT: Generalizable Backdoor Attacks in RLHF via Emotion-Aware Trigger Synthesis31 Aug
- Progressive Behavioral Drift through Compression Valleys in Large Language Models31 Aug
- The Autonomy Tax: Defense Training Breaks LLM Agents31 Aug
- GraftyVul: Synthesising Insecure Programs Through Real-World Vulnerability Grafting31 Aug
- ROPE: Routed Origin Policy Enforcement against Indirect Prompt Injection31 Aug
- ROAST: Risk-aware Outlier-exposure for Adversarial Selective Training of Anomaly Detectors Against Evasion Attacks31 Aug
- eBPF-Based Cybersecurity Mechanisms: A Systematic Literature Review31 Aug
- Removing the Watermark Is Not Enough: Forensic Stealth in Generative-AI Watermark Removal31 Aug
- Password spraying campaign targets AWS root user accounts across 150+ organizations31 Aug
- BengalSEO Part 1: Anatomy of the Operation24 Aug
- The Hugging Face Incident Was a Governance Failure5 Aug
- Security Incident INC-2026-07-28-01 – UK AI Security Institute [pdf]4 Aug
- A First Look at Evo Agentic AppSec: Agentic Remediation and Malicious Code Defense4 Aug
- Inside the keyv npm Compromise: preinstall Malware, Trusted Provenance, and IDE Hooks4 Aug
- Anthropic's Fever Dream: Claude's package that stole real keys31 Jul
- CosmosEscape: Taking Over Every Database in Azure Cosmos DB30 Jul
- Dealing with AI-Generated Extortion30 Jul
- Four incident-response decisions from the Hugging Face breach29 Jul
- The Good, the Bad and the Ugly in Cybersecurity – Week 30 (2026)24 Jul
- LG to Ban Residential Proxies from Smart TV Apps22 Jul
- Breaking Down the White House’s Actions on Post-Quantum Cryptography Readiness2 Jul
- Aikido acquires Root to secure the supply chain30 Jun
- Shipping post-quantum cryptography to Python30 Jun
- The Bear Necessities: A Look at the Drivers, Dynamics, and Applications of the Pro-Russia Influence Ecosystem29 Jun
- The Red Agent POV: Exploiting Broken Object-Level Authorization in an Airline GraphQL API29 Jun
- Snyk VulnBench JS 1.0: Can LLMs Find the Same Bugs Twice?29 Jun
- State Digital Surveillance Risk Landscape17 Jun
- NIST NVD Enrichment Policy Change: Prioritizing Vulnerabilities with Attacker Behavior Signals14 May
- Flash Alert: From Bing Search to Ransomware: Bumblebee and AdaptixC2 Deliver Akira5 Aug
- Version 1.0: Zyxel Firewalls zur Verbreitung von Ransomware missbraucht22 Nov
Updated
24 stories- [UPDATE] [hoch] Red Hat OpenShift: Mehrere Schwachstellen4 Sept
- [UPDATE] [hoch] Red Hat Enterprise Linux (Advanced Cluster Management): Mehrere Schwachstellen4 Sept
- [UPDATE] [hoch] Red Hat Enterprise Linux: Mehrere Schwachstellen4 Sept
- [UPDATE] [mittel] FasterXML Jackson: Schwachstelle ermöglicht Denial of Service4 Sept
- [UPDATE] [mittel] Red Hat Ansible Automation Platform (ansible-core): Schwachstelle ermöglicht Codeausführung4 Sept
- [UPDATE] [hoch] Linux Kernel: Mehrere Schwachstellen3 Sept
- [UPDATE] [mittel] Apache HttpComponents Core: Mehrere Schwachstellen ermöglichen Denial of Service2 Sept
- [UPDATE] [hoch] Internet Systems Consortium BIND: Mehrere Schwachstellen1 Sept
- [UPDATE] [hoch] Red Hat Enterprise Linux (SSSD, libsoup): Mehrere Schwachstellen1 Sept
- Who was behind the attack? Possibly nobody5 Aug
- Of Course We Built a WSUS Ludus Lab5 Aug
- Weaponizing Windows Updates with NotWSUSpicious5 Aug
- A few notes on AWS Nitro Enclaves: KMS integration5 Aug
- “Keep going, bro. You’ve got this!” A data-driven look at how adversaries are weaponizing AI4 Aug
- ConfigManBearPig 2.0 – Things Are Getting Cereal3 Aug
- LLM Heist: Hijacking LiteLLM for Traffic Interception, Key Theft, and Tool-Call Injection3 Aug
- The Xcode Assassin Returns: A Deep Dive Into the Latest XCSSET Version31 Jul
- [tl;dr sec] #335 - Prompt Injection as Role Confusion, PHP Ecosystem Security, New MCP Spec2 Jul
- Google’s Continued Disruption of Malicious Residential Proxy Networks2 Jul
- A Sophisticated Brand Impersonation Scheme: Threats Beneath the Surface1 Jul
- Accelerating EDR Evasion with LLM-Driven Analysis29 Jun
- Jailbreaker: LLM Jailbreak Testing You Can Actually Repeat29 Jun
- [tl;dr sec] #330 - AWS Pathfinding Labs, Running Codex Safely at OpenAI, Glasswing Updates28 May
- Commit to Compromise: A New Threat Actor Targeting the Cryptocurrency Industry's Software Development Infrastructure27 May