AIBearisharXiv – CS AI · Jun 97/10
🧠Researchers demonstrate that popular EEG foundation models (BIOT, LaBraM, EEGPT) leak sensitive neurological attributes despite appearing secure under individual audits. A cross-encoder transfer attack shows that attribute decoders trained on one frozen model successfully transfer to others, indicating shared vulnerabilities that standard defenses like differential privacy fail to adequately address.
AIBullisharXiv – CS AI · Jun 47/10
🧠SharedRequest introduces a privacy-preserving inference framework for large language models that protects user prompt privacy by mixing prompts with noisy variants at the batch level, rather than individual-prompt level. The model-agnostic approach achieves 20% higher utility than differential privacy baselines while reducing query costs by up to 5x, requiring no modifications to LLM architecture.
🧠 ChatGPT
AIBearisharXiv – CS AI · Jun 27/10
🧠Researchers discovered that vision-language models trained on paired chest X-rays and medical reports can re-link de-identified images to their original reports through embedding similarity, creating a privacy vulnerability. The team demonstrated this risk scales with model specialization and developed a differential privacy technique that reduces re-linkage by 62% while preserving diagnostic utility.
AIBearisharXiv – CS AI · Jun 17/10
🧠Researchers propose Bayesian Membership Inference Attacks (BMIA), a novel method that uses Bayesian sampling and Laplace approximation to detect whether specific data points were used in model training. The approach significantly reduces computational overhead compared to existing methods while achieving state-of-the-art attack performance across image, text, and tabular datasets.
AIBullisharXiv – CS AI · Apr 147/10
🧠Researchers propose RPSG, a novel method for generating synthetic data from private text using large language models while maintaining differential privacy protections. The approach uses private seeds and formal privacy mechanisms during candidate selection, achieving high fidelity synthetic data with stronger privacy guarantees than existing methods.
AIBullisharXiv – CS AI · Mar 117/10
🧠Researchers have developed a new framework that enables dataset condensation for non-differentiable clinical AI models like decision trees and Cox regression, using differential privacy to create synthetic medical datasets. This breakthrough allows healthcare institutions to share condensed synthetic data while preserving patient privacy and maintaining model utility across classification and survival prediction tasks.
AIBullishGoogle DeepMind Blog · Oct 237/104
🧠VaultGemma represents a breakthrough as the most capable large language model trained from scratch using differential privacy techniques. This development advances privacy-preserving AI by demonstrating that sophisticated models can be built while maintaining strong data protection guarantees.
AINeutralarXiv – CS AI · Jun 236/10
🧠Researchers introduce π-RAG, a novel retrieval architecture that protects sensitive data in Large Language Models by using the digits of pi as an oblivious indirection layer, eliminating direct exposure of vector embeddings to inversion attacks. The system combines semantic quantization with cryptographic salting to enable privacy-preserving retrieval for compliance-heavy sectors like finance and healthcare.
AINeutralarXiv – CS AI · Jun 236/10
🧠Researchers introduce EPSVec, a differentially-private method for generating synthetic data using large language models that operates significantly more efficiently than existing approaches. By using dataset vectors to steer LLM generation, the technique decouples privacy costs from the number of synthetic samples generated, enabling high-quality synthetic data creation even with limited private datasets.
AINeutralarXiv – CS AI · Jun 116/10
🧠Researchers introduce Robust Privacy (RP), an inference-stage privacy framework that leverages certified robustness principles to prevent adversaries from inferring sensitive attributes or reconstructing training data from model predictions. The approach significantly outperforms differential privacy methods, reducing model inversion attack success rates from 73% to 4% while maintaining 98.4% accuracy, though it remains vulnerable to function-level extraction through model distillation.
AINeutralarXiv – CS AI · Jun 116/10
🧠Researchers developed a federated learning system for ECG anomaly detection that simultaneously achieves GDPR/HIPAA compliance, real-time edge device performance, and clinical-grade detection accuracy across non-uniform hospital data. The system combines differential privacy, quantization, and federated averaging to enable privacy-preserving cardiac monitoring on resource-constrained hardware like Raspberry Pi 4.
AINeutralarXiv – CS AI · Jun 96/10
🧠Researchers evaluated trade-offs between fidelity, privacy, and utility in synthetic image generation across VAE, GAN, and DDPM models under data scarcity conditions. The study reveals that GANs and DDPMs maintain performance better than VAEs when differential privacy mechanisms are applied, suggesting no single generative model excels across all three dimensions simultaneously.
AINeutralarXiv – CS AI · Jun 56/10
🧠Researchers introduce Synapse, a federated learning framework using typed artifacts that enables heterogeneous language models to collaborate without sharing weights or data. The system enables cross-architectural model transfer with minimal performance loss while maintaining formal privacy guarantees and schema-aware merging capabilities.
🧠 GPT-4
AIBullisharXiv – CS AI · Jun 56/10
🧠Researchers introduce InfoShield, a privacy-preserving machine learning technique that maintains depression detection accuracy while preventing the inference of sensitive demographic attributes from speech data. The method uses information-theoretic optimization to reduce mutual information between speech representations and demographic information, addressing a critical barrier to clinical deployment of speech-based mental health screening.
AI × CryptoNeutralarXiv – CS AI · Jun 26/10
🤖Researchers propose SS-ZKR, a privacy-preserving routing protocol that enables multi-agent AI systems to exchange data across organizational boundaries without exposing sensitive information to intermediaries. The protocol combines zero-knowledge proofs, differential privacy, and cryptographic policy compilation to address compliance requirements in regulated industries like finance and healthcare.
AINeutralarXiv – CS AI · Jun 26/10
🧠Researchers introduce Fair Fine-tuning (FFt), a defense mechanism that combines fairness constraints with model fine-tuning to mitigate distribution inference attacks, where adversaries infer sensitive demographic information from machine learning models. The approach reduces adversarial accuracy gaps from ~15% to under 4% across multiple datasets while providing formal theoretical guarantees linking fairness metrics to privacy protection.
🏢 Meta
AINeutralarXiv – CS AI · Jun 16/10
🧠Researchers introduce DPPrefSyn, an algorithm for generating differentially private synthetic preference data to train large language models while protecting user privacy. The method combines the Bradley-Terry preference model with DP-PCA to create synthetic training data from private datasets, achieving competitive alignment performance with formal privacy guarantees.
AINeutralarXiv – CS AI · May 286/10
🧠Researchers introduce mixture mechanisms for differential privacy that combine multiple Gaussian distributions to reduce noise in data queries while maintaining privacy guarantees. These mechanisms substantially outperform existing analytic Gaussian approaches in low-privacy regimes, approaching theoretical optimality with significantly lower noise amplitudes and variances.
AINeutralarXiv – CS AI · May 126/10
🧠diffGHOST is a new conditional diffusion model that synthesizes mobility trajectories while preserving privacy through latent space segmentation. The approach addresses a critical gap in existing generative models that lack formal privacy guarantees despite handling sensitive personal movement data.
AINeutralarXiv – CS AI · May 126/10
🧠Researchers introduce UMEDA, a federated learning framework designed to enable device-free localization across heterogeneous sensors while maintaining privacy. The system uses spectral signal processing and diffusion-based aggregation to align data from different sensor modalities without requiring direct node correspondence, achieving superior performance on multi-modal benchmarks under privacy constraints.
AINeutralarXiv – CS AI · May 116/10
🧠Researchers present the first theoretical framework for differentially private reinforcement learning with general function approximation, achieving regret bounds of Õ(K^3/5) that match linear-case performance. This breakthrough extends privacy guarantees beyond tabular and linear settings, combining batched policy updates with the exponential mechanism for improved privacy-utility tradeoffs in online RL systems.
AINeutralarXiv – CS AI · May 116/10
🧠Researchers propose INO-SGD, a novel algorithm addressing the utility imbalance problem in individualized differential privacy (IDP) machine learning systems. The algorithm strategically down-weights sensitive data batches to prevent underrepresentation of privacy-protected subsets, improving model performance for high-privacy users while maintaining differential privacy guarantees.
AIBullisharXiv – CS AI · May 96/10
🧠PACZero introduces a novel PAC-private fine-tuning mechanism for large language models that achieves usable utility while maintaining zero mutual information leakage, surpassing traditional differential privacy approaches. Using sign quantization of zeroth-order gradients, the method exploits moments of unanimous agreement across candidate subsets to eliminate privacy costs, demonstrating competitive performance on benchmark tasks like SST-2 and SQuAD.
AINeutralarXiv – CS AI · Apr 206/10
🧠Researchers introduce DPrivBench, a benchmark for evaluating how well large language models can reason about differential privacy algorithms and verify their correctness. Testing shows current LLMs handle basic DP mechanisms competently but fail significantly on advanced algorithms, exposing critical gaps in automated privacy reasoning capabilities.
AINeutralarXiv – CS AI · Apr 106/10
🧠Researchers introduce Privacy-Preserving Fine-Tuning (PPFT), a novel training approach that enables LLM services to process user queries without receiving raw text, addressing privacy vulnerabilities in current deployments. The method uses client-side encoders and noise-injected embeddings to maintain competitive model performance while eliminating exposure of sensitive personal, medical, or legal information.
