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

EU Regulatory Changes

1501 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: Multi-Class vs. Multi-Label BERT for CVE-to-CWE Mapping: How Taxonomy Structure Shapes the Errors
arXiv: Beyond Attack-Success Rate: Action-Graded Severity Scale for Tool-Using AI Agents
arXiv: Where to Intervene? Benchmarking Fairness-Aware Learning on Differentially Private Synthetic Tabular Data
arXiv: Mitigating Taint-Style Vulnerabilities in MCP Servers via Security-Aware Tool Descriptions
arXiv: The Blind Curator: How a Biased Judge Silently Disables Skill Retirement in Self-Evolving Agents
arXiv: Beware of Agentic Botnets: Scalable Untargeted Promptware Attacks via Universal and Transferable Adversarial H...
arXiv: How Reliable Is the Multi-Input Heuristic for Bitcoin Address Clustering in Law Enforcement Contexts?
arXiv: Dithered Gaussian Mechanism for Randomness-Efficient Differential Privacy
This paper, published on arXiv, introduces a new mathematical technique called the Dithered Gaussian Mechanism, which aims to make differential privacy more efficient by reducing the amount of rand...
Read analysis →
arXiv: FDIFormer:Protocol-Aware Transformer Learning for False Data Injection Attack Detection in Smart Grid Networks
This publication introduces a novel machine learning model, the FDIFormer, designed to detect False Data Injection Attacks in smart grid networks. The research presents a protocol-aware transformer...
Read analysis →
arXiv: Claimed or Attested? A Commit-Signature Dataset and Identity Trust Tiers across the World of Code
This publication introduces a new dataset and classification framework for assessing the trustworthiness of software code contributions based on digital signatures. The authors propose "Identity Tr...
Read analysis →
arXiv: The Masks We (Think We) Wear: Privacy Threats of Browser-Extension Wallets in the Web3 Ecosystem
This publication, a research paper from arXiv, presents a detailed security analysis of browser-extension wallets used in the Web3 ecosystem, such as those for cryptocurrency and decentralized appl...
Read analysis →
arXiv: Poster: Mind the Gap -- Characterizing the Temporal Blind Spot Between GSB and DNS Resolution
This paper, published on arXiv, identifies a critical timing vulnerability in how internet infrastructure handles domain name resolution, specifically between the Global Server Load Balancer (GSLB)...
Read analysis →
arXiv: Evaluating Fine-Tuning and Metrics for Neural Decompilation of Dart AOT Binaries
This paper, published on arXiv, presents a new method for neural decompilation of Dart AOT (ahead-of-time) binaries, focusing on evaluating fine-tuning techniques and metrics. While not a regulator...
Read analysis →
arXiv: 6G Sensing Security: Distributed Game-Theoretic RL for Urban Beamforming and Attacker Detection
This paper, published on arXiv, proposes a novel technical framework for securing 6G wireless networks, specifically addressing vulnerabilities in urban beamforming and attacker detection. It intro...
Read analysis →
arXiv: A Dual-CRDT Architecture for Decentralized Trust Governance and Evolution
This paper, published on arXiv, proposes a novel technical architecture called Dual-CRDT for managing decentralized trust and governance in AI systems. It is not a regulatory change itself, but a r...
Read analysis →
arXiv: REAN: Reconstruction-aware ECG Anonymization Based on Privacy--Utility Orthogonality
This paper, published on arXiv, introduces a new method called REAN for anonymizing electrocardiogram (ECG) data. It addresses a critical tension in AI safety: the need to protect patient privacy w...
Read analysis →
arXiv: Multi-Channel Spread-Spectrum Code Watermarking
This publication introduces a novel technical method for embedding invisible, tamper-resistant watermarks into AI-generated code using multi-channel spread-spectrum techniques. While not a regulato...
Read analysis →
arXiv: Context-to-Execution Integrity for LLM Agents
arXiv: Bit2Watt: A Cyber-Physical Vulnerability Exploiting GPU Workloads Across Power and Computing Infrastructures
arXiv: ProvICS: A Provenance-based Intrusion Detection for Industrial Control Systems