Latest news
-
AI Security via arXiv cs.CR
DICE: Tilting the Odds through Interface Attackson System One Model
arXiv:2610.14404v1 Announce Type: new Abstract: System One models offer typed judgments that applications can turn into decisions. Their answers, however, are…
-
AI Security via arXiv cs.CR
ByteMark: A Unified Byte-Level Watermark for Detection Across Language Models
arXiv:2610.14427v1 Announce Type: new Abstract: As watermarking of LLM-generated text enters deployment, users increasingly need to check text whose source is unknown…
-
AI Security via arXiv cs.CR
MemForm: Benchmarking Memory Formation Poisoning in LLM Agents
arXiv:2610.14431v1 Announce Type: new Abstract: Long-term memory enables LLM agents to learn from past interactions but exposes memory formation to adversarial…
-
Vulnerabilities via arXiv cs.CR
SecReason: Vulnerability Detection and In-Stream Self-Repair for LLM Code Generation
arXiv:2610.14453v1 Announce Type: new Abstract: Large language models (LLMs) may introduce security vulnerabilities into generated code, which can persist into…
-
Breaches via arXiv cs.CR
Names Are Not Objects: Object-Bound Confinement for Compromised AI Coding Agents
arXiv:2610.14488v1 Announce Type: new Abstract: Coding agents read repositories and web pages that anyone can write, and then act with their user's authority…
-
Breaches via arXiv cs.CR
Decision Uncertainty-Oriented Watermarking (DUOW): Securing Multi-Agent Routing Against Policy Cloning
arXiv:2610.14606v1 Announce Type: new Abstract: Routing policies in Multi-Agent Systems (MAS) represent valuable intellectual property, yet their exposure through…
-
Threat Intel via arXiv cs.CR
Towards Merging-conditional Backdoor Unalignment against Large Language Models
arXiv:2610.14701v1 Announce Type: new Abstract: Model merging has become widely adopted for enhancing the task-specific performance of Large Language Models (LLMs)…
-
Vulnerabilities via arXiv cs.CR
No, Your Tools Are Not Enough: It's Time to Bring Agents to the Hunt
arXiv:2610.14729v1 Announce Type: new Abstract: Modern software systems are large and complex, demanding scalable and autonomous techniques for vulnerability discovery…
- Advertisement
-
Vulnerabilities via arXiv cs.CR
Last Secure Code Benchmark: Is the Code You Vibe Secure?
arXiv:2610.14766v1 Announce Type: new Abstract: AI coding agents now write much of the code that ships, yet whether that code is secure remains unclear. We introduce the…
-
Threat Intel via arXiv cs.CR
When Does Trust Help? A Leakage-Free Evaluation of Representation- and Aggregation-Based Defenses Against Backdoors in Federated Intrusion Detection
arXiv:2610.14767v1 Announce Type: new Abstract: Federated learning (FL) lets distributed intrusion-detection systems (IDS) train on private traffic, but exposes the…
-
AI Security via arXiv cs.CR
Protecting Shared Agent Workflows from Invalidation Attacks with Signed Trajectories and Runtime Verification
arXiv:2610.14795v1 Announce Type: new Abstract: Large language model (LLM)-based agents increasingly rely on shared workflow repositories to transfer experience across…
-
AI Security via arXiv cs.CR
MetaCascade: End-to-End Attacks on Agent Skills via Metadata-Only Poisoning
arXiv:2610.14836v1 Announce Type: new Abstract: As a key extension mechanism for agent systems, Agent Skills package specialized capabilities for reuse by LLM agents. An…
-
Breaches via arXiv cs.CR
StateSync-GKR: Design and Evaluation of a Reusable Sparse GKR Engine
arXiv:2610.14849v1 Announce Type: new Abstract: Authenticated state transitions need proof engines that conserve memory, avoid repeated setup, and preserve request…
-
Vulnerabilities via Ubuntu Security
USN-8925-1: YAML-LibYAML vulnerabilities
Dominique Dumont discovered that YAML-LibYAML did not correctly handle certain reflection operations. If a user or system were tricked into opening a specially crafted…
-
Threat Intel via SANS Internet Storm Center
ISC Stormcast For Wednesday, October 14th, 2026 https://isc.sans.edu/podcastdetail/10136, (Wed, Oct 14th)
(c) SANS Internet Storm Center. https://isc.sans.edu Creative Commons Attribution-Noncommercial 3.0 United States License.
-
AI Security via The Hacker News
OpenAI Employee Arrested for Allegedly Planting Hidden Recorder in Executive Conference Room
A Saudi National and employee of OpenAI has been arrested for allegedly planting a hidden recording device in an executive conference room at the company's San Francisco…
-
Breaches via TechCrunch
Justice Department says North Koreans have infiltrated dozens of U.S. companies
U.S. authorities warn that sanctioned hackers are increasingly targeting Americans to act as facilitators between breached U.S. companies and Pyongyang.
-
Ransomware via ransomware.live
🏴☠️ Dragonforce has just published a new victim : Lane Equipment
Lane Equipment Company has been a trusted food service partner in Texas since 1966, specializing in the sales, leasing, repairs, and parts for commercial kitchen…
-
Vulnerabilities via Splunk
SVD-2026-1006: Advance Notice of Publication of Splunk Security Advisories - October 2026
On October 21, 2026, Splunk PSIRT will publish advisories to disclose security vulnerability advisories for Splunk Enterprise, Universal Forwarder, and SOAR.
-
AI Security via AI Incident Database
OpenAI Pulls Three AI-Generated Math Papers Over Sign Error
TL;DR OpenAI withdrew three of its 722 AI-generated math manuscripts on October 7 after a sign error in a Weil-classes proof collapsed the argument. The flaw cascaded…