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

EU Regulatory Changes

1550 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: VulnGym: Evaluating Vulnerability Management Strategies against Advanced Persistent Threats
arXiv: Near-Tight Theoretical Bounds for Incentive Compatibility in Bitcoin Mining
arXiv: Occluded Oculus: Operationalizing Stylistic Obscurement
arXiv: When LLM Defenses Backfire: Characterizing Safety, Performance, and Cost Trade-offs
arXiv: DeepFaith: Evidence-Grounded LLMs for Faithful Incident Reporting in Multi-Stage APT Defense
arXiv: Quantum-Level Crosstalk Characterization of a 16x16 MEMS Optical Switch for Dynamic Quantum Communications
arXiv: New perspectives for code locality in the rank metric
arXiv: EXE-Bench: Ranking the Tradeoffs of AI-based Windows Malware Detectors for Real-World Usability
arXiv: Just Testing, Move Along: Evasion of LLM-based System Log Interpretation by Prompt Injection
arXiv: Incentives and Market Structure in Intent-Based Exchanges: Evidence from a Solver-Reward Reform
This paper, published on arXiv, analyzes a specific regulatory-style reform within a decentralized exchange system, focusing on how changes to solver rewards in an intent-based trading protocol aff...
Read analysis →
arXiv: SIREN (Luring LLMs onto the Rocks): PAIR-Driven Preference Manipulation in Web-RAG Recommenders
This paper, published on arXiv, presents a novel attack vector called SIREN that exploits how large language models (LLMs) are integrated with web-based retrieval-augmented generation (RAG) systems...
Read analysis →
arXiv: Cleaning the NTP Pool: Detecting and Mitigating NTP-Sourced IPv6 Scanning
This paper, published on arXiv, presents a technical analysis of how the Network Time Protocol (NTP) pool can be exploited to conduct large-scale IPv6 scanning, effectively mapping active IPv6 addr...
Read analysis →
arXiv: ISPCloak: Weaponizing ISP for Optimization-Free Physical Camouflage against Deepfake Detectors
This paper, published on arXiv, introduces a novel adversarial attack called ISPCloak that weaponizes the image signal processing (ISP) pipeline—the standard hardware and software chain in cameras—...
Read analysis →
arXiv: PrivDNN: A Secure Multi-Party Computation Framework for Deep Learning using Partial DNN Encryption
This publication introduces PrivDNN, a technical framework for secure multi-party computation in deep learning that uses partial encryption of deep neural networks. While not a regulatory change it...
Read analysis →
arXiv: No Snake Oil: Verifying Python Package Builds
This paper, published on arXiv, presents a technical framework for verifying the integrity of Python package builds, addressing a critical gap in software supply chain security. It proposes methods...
Read analysis →
arXiv: Decentralized Compute on Untrusted Hardware Using Intel TDX and Encrypted CVMs
This paper, published on arXiv, proposes a technical architecture for running decentralized compute workloads on untrusted hardware using Intel TDX and encrypted confidential virtual machines. Whil...
Read analysis →
arXiv: Certified in Theory, Broken in Practice: Assumption Gaps in Cryptographic Model Certification
This paper, published on arXiv in July 2026, presents a critical analysis of cryptographic model certification within the AI safety framework. It argues that current certification processes for cry...
Read analysis →
arXiv: ToolGuardian: Declarative Security for AI Agent-Tool Interactions
This publication introduces ToolGuardian, a declarative security framework designed to govern how AI agents interact with external tools and APIs. It proposes a structured approach to defining and ...
Read analysis →
arXiv: A Maximum Entropy Implementation of Differential Privacy Under Linear Invariants
arXiv: Dynamic Capability Scoping for Enterprise AI Agents: A Synthetic Dataset and Three-Source Permission Architecture