Currently free during beta - premium features coming soon. Subscribe now to lock in early access.
AI_SAFETY

EU Regulatory Changes

1574 changes tracked across 24 compliance frameworks including DORA, NIS2, GDPR, EU AI Act, Cyber Resilience Act, and more.

All DORA NIS2 GDPR CSRD MaRisk ISO27001 EU_AI_ACT CRA DSA DMA eIDAS2 SOC2 PCI_DSS HIPAA ISO42001 AMLD6 PSD3 DATA_ACT GPSR CER EUDR CVE BREACH AI_SAFETY
arXiv: Knowledge Over Parameters: Evolving Smart Contract Vulnerability Detection
arXiv: Trust Boundary Semantic Gaps: A Multi-dimensional Analysis and Mitigation for Security-by-Design
arXiv: Pmeta-TLA: Backdoor Attacks for Speech Classification Models via Meta-Learning with Timbre Leakage Attack
arXiv: Beyond Gradient-Based Attacks: Adversarial Robustness and Explainability Stability in Cybersecurity Classifiers
arXiv: VeriChat: An Agentic Conversational AI Assistant for Hardware Security Verification
arXiv: AgentFlow: Building Agent Dependency Graphs for Static Analysis of Agent Programs
arXiv: LIB-TRAP: Standard Cell Library Hardware Trojan Risk Assessment and Prevention
arXiv: Overthink-Triggered Slowdown Attacks on LVLM-Based Robotic Systems
arXiv: Janus: a Playground for User-Involved Agentic Permission Management
arXiv: Unveiling the Non-Monotonic Effect of Privacy on Generalization under Byzantine Robustness
arXiv: Hamm-Grams: An Algorithm for Mining Regular Expressions of Bytes
arXiv: From Forgeries to Foundation Models: A Systematic Survey of Identity Document Attack and Detection
arXiv: The Rise and Fall of Google's Privacy Sandbox
A new academic paper published on arXiv, titled "The Rise and Fall of Google's Privacy Sandbox," provides a critical retrospective analysis of Google's initiative to phase out third-party cookies i...
Read analysis →
arXiv: High-Performance NTT Accelerators for PQC leveraging Unified Redundant Arithmetic and Fine-Tuned Microarchitec...
This publication from arXiv, dated July 1, 2026, presents a technical paper detailing new hardware accelerators for Post-Quantum Cryptography (PQC). The paper describes a method to significantly sp...
Read analysis →
arXiv: Safe Alone, Unsafe Together: Safeguarding Against Implicit Toxicity When Benign Images Combine
This publication, a pre-print from arXiv dated July 2026, presents a novel vulnerability in multimodal AI systems. It demonstrates that individual benign images, when processed together by a model,...
Read analysis →
arXiv: HARC: Coupling Harmfulness and Refusal Directions for Robust Safety Alignment
This paper, published on arXiv, introduces a new technical framework called HARC, which addresses a critical vulnerability in large language models (LLMs). The research demonstrates that current sa...
Read analysis →
arXiv: Cross-Domain Generalization Failure in Lightweight Intrusion Detection Models for IIoT Networks
A new preprint from arXiv, published on July 1, 2026, presents research demonstrating that lightweight intrusion detection models used in Industrial Internet of Things (IIoT) networks suffer from s...
Read analysis →
arXiv: Beyond the Prompt: Jailbreaking Function-Calling LLMs via Simulated Moderation Traces
This paper, published on arXiv, details a novel vulnerability in large language models (LLMs) that use function-calling capabilities. The research demonstrates that attackers can bypass safety guar...
Read analysis →
arXiv: Minos: A Multi-Agent Collaborative Framework for Provenance-Based Backward Tracking
A new research paper titled "Minos: A Multi-Agent Collaborative Framework for Provenance-Based Backward Tracking" has been published on arXiv, proposing a technical framework for tracing the origin...
Read analysis →
arXiv: KidnapRAG: A Black-Box Attack for Hijacking Reasoning in Agentic Retrieval-Augmented Generation Systems
A new research paper, KidnapRAG, published on arXiv, details a novel black-box attack targeting agentic Retrieval-Augmented Generation (RAG) systems. This attack demonstrates how malicious actors c...
Read analysis →