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

EU Regulatory Changes

668 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: Calibration Without Comprehension: Diagnosing the Limits of Fine-Tuning LLMs for Vulnerability Detection in Sy...
A new research paper published on arXiv, titled "Calibration Without Comprehension: Diagnosing the Limits of Fine-Tuning LLMs for Vulnerability Detection in Systems Software," raises significant co...
Read analysis →
arXiv: A-COMPASS: Formal Foundations for Anonymity Analysis in Microdata
This publication introduces A-COMPASS, a formal mathematical framework for analyzing anonymity in microdata, which is detailed, individual-level data often used in research and analytics. The paper...
Read analysis →
arXiv: Analyzing Defensive Misdirection Against Model-Guided Automated Attacks on Agentic AI Systems
This paper, published on arXiv, presents a new analysis of defensive techniques against automated attacks on agentic AI systems—AI that can autonomously take actions. It specifically examines how "...
Read analysis →
arXiv: Image Encryption Algorithm Based on Convolutional Neural Networks and Dynamic S-Box Generation
This publication from arXiv presents a novel image encryption algorithm that integrates convolutional neural networks with dynamic S-box generation. While not a regulatory change itself, it signals...
Read analysis →
arXiv: Multi-View Decompilation for LLM-Based Malware Classification
This paper, published on arXiv, presents a novel technical approach for classifying malware using large language models (LLMs) through a process called multi-view decompilation. Rather than a regul...
Read analysis →
arXiv: LLM agent safety, multi-turn red-teaming, jailbreak benchmarks, adversarial robustness, safety-critical systems
This paper, published on arXiv, presents a new framework for evaluating the safety of large language model (LLM) agents, specifically focusing on "multi-turn red-teaming" and adversarial robustness...
Read analysis →
arXiv: bioETH-Beacon: A Confidential On-Chain Genomic Beacon with Encrypted Counts, Filters, and Bounded Noise over a...
This publication introduces bioETH-Beacon, a technical framework for running genomic data queries on a blockchain while preserving patient confidentiality. It uses a fully homomorphic encryption sc...
Read analysis →
arXiv: Quantization as a Malicious Task: Removing Quantization-Conditioned Backdoors via Task Arithmetic
arXiv: TrustMix: How to Mix Messages in a Mobile Ad-hoc Network
arXiv: GNSS Spoofing Threat for V2X communications
arXiv: Accelerating Trust Convergence in IIoT: A ML Approach for Dynamic Network Conditions
arXiv: Artificial Intelligence as Game Changer in Cybersecurity: What We Learned in 2025-2026, and how this is releva...
arXiv: A Measurement Study of Cryptographic Misuse in Embodied AI Mobile Applications
arXiv: AutoTam: Specifying Secure Protocol Implementations with Tamarin Model Generation
arXiv: FFinRED: An Expert-Guided Benchmark Generation and Evaluation Framework for Financial LLM Red-Teaming
arXiv: Low-Cost Multi-Precision Systolic Arrays for Accelerating FHE NTTs on AI ASICs
arXiv: Heterogeneous LLM Debate Under Adversarial Peers: Honest Gains, Replacement Costs, and Resilience
arXiv: DISARM: Target Electronic Device Informed Mitigation of Software Runtime Side-Channel Vulnerabilities
arXiv: SafeSpec: Fast and Safe LLM via Dynamic Reflective Sampling
arXiv: When Global Gating Is Enough: Admission-Time Hubness Control in Anisotropic Vector Retrieval Systems