341 articles tagged with #privacy. AI-curated summaries with sentiment analysis and key takeaways from 50+ sources.
Researchers propose a new framework for 'selective forgetting' in Large Reasoning Models (LRMs) that can remove sensitive information from AI training data while preserving general reasoning capabilities. The method uses retrieval-augmented generation to identify and replace problematic reasoning segments with benign placeholders, addressing privacy and copyright concerns in AI systems.
ETHGlobal Cannes 2026 announced 10 finalists featuring projects focused on AI agents, privacy infrastructure, and on-chain prediction markets. Notable projects include ENShell, DIVE, Corpus, and VEIL VPN, representing some of the most technically advanced submissions in ETHGlobal's history.
The Cardano community is criticizing the architecture of Midnight, Cardano's privacy-focused sidechain project. The criticism has prompted Cardano founder Charles Hoskinson to publicly respond and defend the project.
Researchers introduce VLM-UnBench, the first benchmark for evaluating training-free visual concept unlearning in Vision Language Models. The study reveals that realistic prompts fail to genuinely remove sensitive or copyrighted visual concepts, with meaningful suppression only occurring under oracle conditions that explicitly disclose target concepts.
Researchers successfully fine-tuned LLaMA 3.1-8B for medical transcription in Finnish, a low-resource language, achieving strong semantic similarity despite low n-gram overlap. The study used simulated clinical conversations from students and demonstrates the feasibility of privacy-oriented domain-specific language models for clinical documentation in underrepresented languages.
Researchers developed lightweight generative AI models for creating synthetic network traffic data to address privacy concerns and data scarcity in network traffic classification. The models achieved up to 87% F1-score when classifiers were trained solely on synthetic data, with transformer-based approaches providing the best balance of accuracy and computational efficiency.
Researchers have developed TAAC, a framework for trustable audio-based depression diagnosis that protects user identity information while maintaining diagnostic accuracy. The system uses adversarial loss-based subspace decomposition to separate depression features from sensitive identity data, enabling secure AI-powered mental health screening.
T-REX Network and Zama have partnered to launch institutional-grade confidentiality infrastructure specifically designed for Real World Asset (RWA) tokenization. This collaboration aims to provide enhanced privacy and security features for institutional clients looking to tokenize traditional assets on blockchain networks.
