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

EU Regulatory Changes

371 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: Benchmarking Empirical Privacy Protection for Adaptations of Large Language Models
arXiv: Brain-Prompt Injection: A Route-Safety Audit for BCI-LLM Agents
arXiv: Trustworthy Smart Fabs via Professional Proxies: Scaling Safe and Sustainable by Design (SSbD) through Industr...
arXiv: The Injection Paradox: Brand-Level Suppression in Safety-Trained LLM Recommendations via RAG Context Injection
arXiv: Empirical Evaluation of Large Language Models for Migration of Code Fragments to Post-Quantum Cryptography
This publication presents an empirical evaluation of large language models (LLMs) for automatically migrating existing code fragments to post-quantum cryptography (PQC) algorithms. The study assess...
Read analysis →
arXiv: Defending Jailbreak Attacks on Large Language Models via Manifold Trajectory Kinetics
This paper, published on arXiv, introduces a novel technical method called Manifold Trajectory Kinetics designed to defend large language models against "jailbreak" attacks—prompts that trick AI sy...
Read analysis →
arXiv: Authorized and Verifiable Searchable Encryption Based on Public Key Equality Test for Cloud Storage
This document is a research paper proposing a new cryptographic method for cloud storage, not a formal regulatory change. It introduces an "Authorized and Verifiable Searchable Encryption" scheme b...
Read analysis →
arXiv: Rethinking IoT Intrusion Detection: Augmenting Routing Metrics with Radio Features
This publication, dated June 5, 2026, presents a novel framework for intrusion detection in Internet of Things (IoT) networks. The core change is a proposed methodology that moves beyond traditiona...
Read analysis →
arXiv: The Capacity of Information-Theoretic Secure Aggregation in Federated Learning
This publication from arXiv presents a theoretical analysis of the capacity limits for information-theoretic secure aggregation in federated learning. It does not introduce a new regulation or bind...
Read analysis →
arXiv: A Large-Scale Per-Speaker Analysis of Re-identification Risk in Speech Anonymization
This publication from June 2026 presents a large-scale study on the re-identification risk of speech anonymization techniques, specifically analyzing how well current methods protect individual spe...
Read analysis →
arXiv: Synthetic APTs: the Collapse of TTP-Based Attribution
A new preprint from arXiv, titled "Synthetic APTs: the Collapse of TTP-Based Attribution," published on June 5, 2026, presents a significant challenge to existing cybersecurity threat intelligence ...
Read analysis →
arXiv: From Privacy to Workflow Integrity: Communication-Graph Metadata in Autonomous Agent Interoperability
This paper, published on arXiv under the AI_SAFETY framework, presents a novel analysis of communication-graph metadata risks in autonomous agent systems. It argues that current privacy and safety ...
Read analysis →
arXiv: MalSkillBench: A Runtime-Verified Benchmark of Malicious Agent Skills
A new research paper, MalSkillBench, has been published on arXiv, presenting a benchmark designed to evaluate the capabilities of AI agents in performing malicious cyber tasks. The framework system...
Read analysis →
arXiv: TRACE: Trajectory Reasoning through Adaptive Cross-Step Evidence Aggregation for LLM Agents
This paper, published on arXiv, introduces TRACE, a new reasoning framework for large language model agents. TRACE improves how AI systems handle complex, multi-step tasks by aggregating evidence a...
Read analysis →
arXiv: Fast Bounded-Independence Functions and Their Duals
arXiv: The Sound of Malware: A Memory Forensics Approach for Android Malware Analysis via Audio Signals
arXiv: HAVE: Host Active Verification Engine for Closing the Contextual Reality Gap in Security Digital Twins
arXiv: DPAgent-in-the-Middle: Agentic Defense and Repair Against AI-Groomed Deceptive Patterns
arXiv: Blockchain Infrastructure for Intelligent Cyber--Physical--Social Systems:Post-Quantum Security, Interoperabil...
arXiv: FDM: A Framework for Decision-making to build ML-based Malware detection systems