Latest news
-
AI Security via arXiv cs.CR
Permitted but Unauthorized: Affordance-Induced Delegated-Scope Violations in Multi-Agent LLM Systems
arXiv:2611.04086v1 Announce Type: new Abstract: Multi-agent LLM systems place several agents on shared infrastructure, where each agent holds permissions far broader…
-
AI Security via arXiv cs.CR
ScrambleMark: Zero-Query Watermark-Agnostic Black-Box Evasion of Modern Speech Watermarking
arXiv:2611.04122v1 Announce Type: new Abstract: Speech watermarking is becoming a key defense for digital copyright and content authentication. Existing attacks treat…
-
Breaches via arXiv cs.CR
VerfGraph: Efficient Zero-Knowledge Verifiable Graph Query Evaluation System
arXiv:2611.04240v1 Announce Type: new Abstract: Graph databases are often used to manage relationship-rich data, whose graph structure and attributes may be sensitive…
-
Vulnerabilities via arXiv cs.CR
When Debate Amplifies Risk: Social-Bias Vulnerabilities in Multi-Agent Safety Deliberation
arXiv:2611.04289v1 Announce Type: new Abstract: Retrieval-augmented Web applications increasingly employ multi-agent debate (MAD) to aggregate conflicting, noisy, and…
-
Threat Intel via arXiv cs.CR
The Native Menace: Host-to-Browser Cross-Boundary Attacks via Native Messaging
arXiv:2611.04291v1 Announce Type: new Abstract: Modern browsers are complex software platforms that users can customize through browser extensions to further expand…
-
AI Security via arXiv cs.CR
Can Benign Skills Induce Misaligned Behavior in LLM Agents?
arXiv:2611.04299v1 Announce Type: new Abstract: LLM agents increasingly rely on reusable skills to plan tasks, invoke tools, recover from errors, and complete workflows…
-
AI Security via arXiv cs.CR
Bilateral Protection for Private LoRA Inference with Intel TDX
arXiv:2611.04311v1 Announce Type: new Abstract: Cloud services using private low-rank adaptation (LoRA) must protect client inputs and outputs alongside owners'…
-
AI Security via arXiv cs.CR
CloakFEM: Agentic FEM Code Generation for Red-Teaming Computational Resource Exhaustion
arXiv:2611.04356v1 Announce Type: new Abstract: Finite element method (FEM) solvers are foundational to scientific and engineering simulation, yet their susceptibility…
- Advertisement
-
AI Security via arXiv cs.CR
Voluntary Servitude: Weaponized Safety Warning Mechanisms for Controlling AI Agents
arXiv:2611.04393v1 Announce Type: new Abstract: LLM agents increasingly act on information retrieved from public sources they treat as credible, which turns a…
-
Threat Intel via arXiv cs.CR
TxBench: A Benchmark of Attack Transactions and Evidence-Backed Negatives on Seven EVM Chains
arXiv:2611.04411v1 Announce Type: new Abstract: Researchers evaluate detectors of malicious smart-contract transactions on datasets that each paper assembles anew. Their…
-
Threat Intel via arXiv cs.CR
A Security Analysis of Verifiable Credential Verification Pipelines: Attack Surface Arising from JSON-LD/RDFC Canonicalization Compared with SD-JWT VC and mdoc
arXiv:2611.04461v1 Announce Type: new Abstract: We evaluate, at implementation level and under an explicit threat model, the verification pipelines of three Verifiable…
-
AI Security via arXiv cs.CR
Perseus: A Full-Stack System for Homomorphic Model Inference
arXiv:2611.04488v1 Announce Type: new Abstract: Large Language Models (LLMs) have made it exceptionally convenient for users to send their private data to model…
-
AI Security via arXiv cs.CR
The Devil is in the Context: Nine Dimensions of Prompt Injection Attacks
arXiv:2611.04672v1 Announce Type: new Abstract: Prompt injection is usually discussed and evaluated as a single, homogeneous type of attack. However, vulnerabilities…
-
AI Security via arXiv cs.CR
Right Tool, Wrong Parameters: Defending LLM Agents Against Parameter Poisoning
arXiv:2611.04673v1 Announce Type: new Abstract: Indirect Prompt Injection (IPI) poses a critical threat to LLM agents. Existing defenses largely focus on whether an…
-
Vulnerabilities via arXiv cs.CR
BrainwashedLLM: Jailbreaking via Automated Adversarial Chain of Thoughts Discovery
arXiv:2611.04717v1 Announce Type: new Abstract: Large Reasoning Models (LRMs) achieve strong performance on complex tasks by generating explicit Chain-of-Thought (CoT)…
-
AI Security via Help Net Security
Anthropic open-sources FlashProve, its provable inference prototype for MoE models
Anthropic’s Frontier Red Team, with collaborators at EPFL, MIT and StarkWare, built FlashProve, a prototype that produces a cryptographic proof that one specific…
-
Threat Intel via SANS Internet Storm Center
ISC Stormcast For Wednesday, November 4th, 2026 https://isc.sans.edu/podcastdetail/10164, (Wed, Nov 4th)
(c) SANS Internet Storm Center. https://isc.sans.edu Creative Commons Attribution-Noncommercial 3.0 United States License.
-
Breaches via Dark Reading
Philippine Migrant Worker Data Likely Leaked in 69GB Government Breach
An unknown operator appears to have exfiltrated SQL Server backups from the Philippine Overseas Employment Administration in under a half hour.
-
Threat Intel via DataBreaches.net
French blood storage units targeted in cyberattack
OMM reports: Cold-storage facilities used by France’s national blood service were targeted in a cyberattack, with units containing blood samples reportedly programmed to…
-
Breaches via ransomware.live
🏴☠️ Byod has just published a new victim : IHG Hotels & Resorts (Estate)
Nice LDAP Binding Password being '123'. Rotating things won't really save you casue it's not hard to hop back on your network. Anyways, we now have; Customer Data, AWS…