Advance Zero Trust for AI: New tools and guidance to secure AI agents and DevSecOps
Explorar
Noticias de IA
1359 elementos — filtrados, clasificados y sin duplicados
Meta Ran Ads That Contained AI-Generated Child Sexual Abuse Imagery
Researchers watched OpenAI, Anthropic models take extreme measures in hacking test
Cybersecurity Concerns After OpenAI, Anthropic Tests
Rogue AI agents created fake online identities in another hacking attempt
OpenAI, Anthropic Model Tests Reveal More ‘Unsanctioned’ Actions
OpenAI and Anthropic models went on a hacking spree when tested by the UK's AI research i…
DiagChain: A Diagnostic Benchmark for Evaluating LLM Agents on Evidence-Grounded Attack C…
Steganalysis of Adaptive Covert Collusion in Tool-Using Agent Populations: A Black-Box, C…
Privacy-Preserving AI Verification via Minimal Information Disclosure
AI Security Leaderboard: Methodology, Results and Minimal Standard
Moving the Safety Barrier: Dynamic Routing Adaptive Alignment Against White-Box Attacks
Security-First Evaluation of Text-to-Terraform: Benchmarking LLMs and SLMs for Secure IaC…
A Security-Oriented Lifecycle Model for Large Language Model Systems
DenialRAG: Single-Document RAG Poisoning via Embedded Parametric Denial
AgenticSCR: An Autonomous Agentic Secure Code Review for Immature Vulnerabilities Detecti…
Attribute-based Undetectable Watermarking for Generative AI Models
Single Canonical Prompts Underestimate LLM Safety's Surface-Form Sensitivity
Malice in Agentland: Down the Rabbit Hole of Backdoors in the AI Supply Chain
FakeI2V-Bench: Benchmarking the Applicability of Image-level Deepfake Detectors for Deepf…
MissClick: Exploiting Digit-Serialized Coordinates to Attack GUI Grounding Models
Risky Business: Measuring The Faithfulness-Safety Tension
ADMITBench: A Safety-Governed Reference Framework for Evaluating the Admissibility of Ind…
$S^3$: Improving Agent Safety through Multi-Stage Defense
Toward Understanding the Transferability of Adversarial Suffixes in Large Language Models
LatentGuard: Efficient and Inspectable Latent Reasoning for LLM Safeguards
Your Agentic LLMs Secretly Encode Latent Signals of Indirect Prompt-Injection Exposure
Permission Denied: Policy-Graded Evaluation of Coding Agents in Hardened Environments
MAFIA: Query-Only Memory Attacks via Probing and Factual Injection against Audited LLM Ag…
A Blind Spot in Alignment: Quantifying Biosecurity Risks in Large Language Models