External signal
WorldMark: A Plug-and-Play World Knowledge Interface for Cross-Host Language Model Watermarking (opens external site in a new tab)
arXiv CS.CR
Inteligencia externa
Atlas News Radar
Señales externas recientes sobre seguridad de IA, sistemas agénticos y gobierno — recopiladas de arXiv, BleepingComputer, SecurityWeek, Schneier, Google AI y OWASP GenAI Security.
External signal
Fairis: Fairness-Aware Aggregation with Provable Influence Containment against Fairness Poisoning Attacks in Collaborati (opens external site in a new tab)
arXiv CS.CR
External signal
CyberForge: Verified Vulnerability Injection at Repository Level for Cybersecurity Agent Training (opens external site in a new tab)
arXiv CS.CR
External signal
StepJack: Benchmarking Computer-Use Agent Safety Against Multi-Step Indirect Prompt Injection (opens external site in a new tab)
arXiv CS.CR
External signal
Canonicalization Failures as a Recurring Vulnerability Class: Representation Divergence in Cryptographic Systems and Its (opens external site in a new tab)
arXiv CS.CR
External signal
Model Confidence Under Answer-Preserving Attacks: An Informativeness-Manipulability Frontier (opens external site in a new tab)
arXiv CS.CR
External signal
WhiteNet: Robust Identification of Overlapping IEEE 802.11 Signals Across Unseen Channels (opens external site in a new tab)
arXiv CS.CR
External signal
From Documentation to Zero-day Vulnerabilities: LLM-Driven Fuzzing of JavaScript Engines in PDF Readers (opens external site in a new tab)
arXiv CS.CR
External signal
CyberLLM: A Multi-Agent LLM Framework for Autonomous Detection and Guarded Response in Automotive Cybersecurity (opens external site in a new tab)
arXiv CS.CR
External signal
Policy-Masked Private Experts: Auditable and Reversible Capability Access Control in Sparse MoE Models (opens external site in a new tab)
arXiv CS.CR
External signal
CISA: SonicWall SMA1000 flaws now exploited by ransomware gangs (opens external site in a new tab)
BleepingComputer
External signal
When Credentials Are No Longer Enough: Device Trust in the AI Era (opens external site in a new tab)
BleepingComputer
External signal
Member of The Com sent to prison for blackmail, sextortion (opens external site in a new tab)
BleepingComputer
External signal
LexisNexis shuts down services after suspicious activity on servers (opens external site in a new tab)
BleepingComputer
External signal
Valve notifies Steam hardware customers of a data breach (opens external site in a new tab)
BleepingComputer
External signal
Critical Progress LoadMaster flaw now actively exploited in attacks (opens external site in a new tab)
BleepingComputer
External signal
OpenAI’s Upcoming Astra Model Raises Autonomous Cyberattack Concerns (opens external site in a new tab)
SecurityWeek
External signal
Stealthium Targets Security Blind Spots in AI Accelerators and Neo-Clouds (opens external site in a new tab)
SecurityWeek
External signal
Cisco Warns of High-Severity ClamAV Vulnerabilities With Public PoC (opens external site in a new tab)
SecurityWeek
External signal
‘Ghostjacking’ Attack Uses Poisoned Logs to Turn AI Agents Bad (opens external site in a new tab)
SecurityWeek
External signal
New Jersey, Alabama Join States Targeted in Water Cyberattacks (opens external site in a new tab)
SecurityWeek
External signal
Metabase Patches Vulnerability Exploited as Zero-Day (opens external site in a new tab)
SecurityWeek
External signal
Novel Private APN Pivot Let Hackers Sabotage Second Polish Energy Facility (opens external site in a new tab)
SecurityWeek
External signal
CISA Urges Immediate Patching of Exploited Progress LoadMaster Vulnerability (opens external site in a new tab)
SecurityWeek
External signal
Corporate Data Stolen in Levi Strauss Cyberattack (opens external site in a new tab)
SecurityWeek
External signal
Critical Flaws Discovered in Belgian eID Software Used by 2 Million People (opens external site in a new tab)
SecurityWeek
External signal
Hackers breach TrueConf to trojanize client installers with backdoors (opens external site in a new tab)
BleepingComputer
External signal
Metabase SQLi zero-day exploited in customer data-theft attacks (opens external site in a new tab)
BleepingComputer
External signal
Unlimited Technology Systems breach impacts 3.8 million people (opens external site in a new tab)
BleepingComputer
External signal
Levi Strauss & Co. says hackers stole corporate data in cyberattack (opens external site in a new tab)
BleepingComputer
External signal
Twin Agent: Context Residual Compression for Privilege Separated Agents (opens external site in a new tab)
arXiv Cryptography and Security
llm appsecagentsidentity permissions
External signal
ChainWatch: A Kill Chain-Aligned Sequential Detection Framework for Multi-Step Attacks in MCP-Based AI Agent Systems (opens external site in a new tab)
arXiv Cryptography and Security
llm appsecagents
External signal
The Chronos Vulnerability: A Taxonomy of Temporal Persistence and Memory-Based Deception in Agentic AI (opens external site in a new tab)
arXiv Cryptography and Security
agentssecurity general
External signal
Know Your Agent: Reconnaissance-Driven Pentesting of AI Agents (opens external site in a new tab)
arXiv Cryptography and Security
llm appsecagentssecurity general
External signal
How enterprise GenAI can amplify ransomware risk — and how to contain it (opens external site in a new tab)
BleepingComputer
agentsidentity permissions
External signal
Microsoft Azure DevOps MCP Flaw Lets Hidden PR Comments Hijack AI Review Agents (opens external site in a new tab)
The Hacker News
llm appsecagents
External signal
The Balkanization of Execution-Security Research for AI Coding Agents: Isolation, Access Control, and Time-of-Check-to-Time-of-Use Vulnerabilities (opens external site in a new tab)
arXiv Cryptography and Security
llm appsecagentsidentity permissions
External signal
Beyond the Leaderboard: A Synthesis of Tool-Use, Planning, and Reasoning Failures in Large Language Model Agents (opens external site in a new tab)
arXiv Artificial Intelligence
llm appsecagents
External signal
PolyWorkBench: Benchmarking Multilingual Long-Horizon LLM Agents (opens external site in a new tab)
arXiv Artificial Intelligence
llm appsecagents
External signal
Unicode TAG-Block Concealment of Tool-Metadata Payloads in the Model Context Protocol: An Approval-View Fidelity Gap Across Three Independent Server Implementations (opens external site in a new tab)
arXiv Cryptography and Security
llm appsecagents
External signal
Information Gain-based Rollout Policy Optimization: An Adaptive Tree-Structured Rollout Approach for Multi-Turn LLM Agents (opens external site in a new tab)
arXiv Artificial Intelligence
llm appsecagents
External signal
Beyond Static Evaluation: Building Simulation Environments for Scalable Agentic Reinforcement Learning (opens external site in a new tab)
arXiv Artificial Intelligence
agents
External signal
TurnOPD: Making On-Policy Distillation Turn-Aware for Efficient Long-Horizon Agent Training (opens external site in a new tab)
arXiv Artificial Intelligence
agents
External signal
aiAuthZ: Off-Host, Identity-Bound Authorization for AI Agents (opens external site in a new tab)
arXiv Cryptography and Security
agentsidentity permissions
External signal
Prompt-to-Paper: Agentic AI System for Bioinformatics (opens external site in a new tab)
arXiv Artificial Intelligence
agentssupply chain
External signal
Expanding Managed Agents in Gemini API: background tasks, remote MCP and more (opens external site in a new tab)
Google AI Blog
llm appsecagents
External signal
Inside Elastic InfoSec's agentic SOC: cutting alert triage from 30 minutes to under 3 (opens external site in a new tab)
Elastic Security Labs
agents
External signal
‘BioShocking’ Attack Tricks AI Browsers Into Stealing Credentials (opens external site in a new tab)
SecurityWeek
agents
External signal
Snyk VulnBench JS 1.0: Can LLMs Find the Same Bugs Twice? (opens external site in a new tab)
Snyk Blog
ai securityllm appsecsecurity general
External signal
One Million Passports Leaked Online (opens external site in a new tab)
Schneier on Security
identity permissions
External signal
Interesting Paper Exploring Prompt Injection (opens external site in a new tab)
Schneier on Security
llm appsec
External signal
From vulnerability report to CVE draft in minutes: how Elastic automated security advisories with AI (opens external site in a new tab)
Elastic Security Labs
llm appsecagentssecurity general
External signal
What nearly 10,000 developer environments reveal about agentic development risk (opens external site in a new tab)
Snyk Blog
agentsidentity permissionssupply chain
External signal
macOS.Gaslight | Rust Backdoor Turns Prompt Injection on the Analyst, Not the Sandbox (opens external site in a new tab)
SentinelOne Labs
llm appsecidentity permissions