DeepMind Blog
Oct 23, 2025
VaultGemma: The world's most capable differentially private LLM
We introduce VaultGemma, the most capable model trained from scratch with differential privacy.
Creating Privacy Preserving AI with Substra
Related stories
arXiv Machine Learning
Oct 2
Privacy in Personalized AI Is a System Property, Not Just a Model Property
The paper argues that privacy in personalized AI should be viewed as a system-level issue rather than just a model-level one. It identifies four interconnected privacy‑risk channels in personalized AI and proposes four system‑level requirements—interaction trajectories, internal information flows, indirect leakage, and the privacy‑utility trade‑off—for evaluating privacy. The authors call for these requirements to be systematically incorporated into privacy audits of personalized AI systems.
arXiv AI
Jul 9
Security and Privacy in Agentic AI: Grand Challenges and Future Directions
arXiv:2607. 06608v1 Announce Type: cross Abstract: We present key challenges and future research directions in the security and privacy of agentic AI, based on a horizon-scanning exercise that brought together thirty leading international experts from academia, industry, and government to engage in focused discussions and collaborative exercises on the emerging risks associated with the growing agency of AI.
arXiv Machine Learning
Aug 31
Performative Privacy: When Differential Privacy Maximizes Utility
arXiv:2608.28198v1 Announce Type: new Abstract: Privacy-preserving learning is often motivated by the idea that protecting users' data can preserve trust and thus participation, improving utility in...
arXiv Machine Learning
Sep 1
Revisiting the Provable-Auditable Privacy Gap of DP-SGD
arXiv:2608.28934v1 Announce Type: new Abstract: Differential privacy (DP) has traditionally been used to provide theoretical upper bounds on an algorithm's stability to changing its training data. In...
arXiv Machine Learning
Jun 4
DPDL: Towards Differential Privacy Preservation in Decentralized Stochastic Learning on Non-IID Data
arXiv:2606. 04399v1 Announce Type: new Abstract: In the paradigm of decentralized learning, a group of agents collaborate to train a global model using distributed datasets without a central server.
arXiv AI
Jul 13
How to DP-fy Your Data: A Practical Guide to Generating Synthetic Data With Differential Privacy
arXiv:2512. 03238v2 Announce Type: replace-cross Abstract: High quality data is needed to unlock the full potential of AI for end users.
arXiv AI
Oct 5
HXAI: Hierarchical Privacy-Preserving Explainable AI in Distributed Energy Systems
arXiv:2610.02504v1 Announce Type: new Abstract: Balancing electricity demand and supply is increasingly difficult due to the inherent intermittency of renewable power generation and the stochastic po...
DeepMind Blog
Sep 23
Advancing Private AI Compute with secure, server-side memory
DeepMind introduces private, server‑side memory to its Private AI Compute platform, enabling personal AI applications to operate with secure, isolated memory on the server side. This enhancement focuses on protecting user data while allowing AI models to run efficiently in a private environment. The update expands the capabilities of Private AI Compute by adding a new layer of security and privacy for personal AI use cases.
Hugging Face Trending Papers
4d ago
Corralling AI Risks with Grassroots Contracts
AI's inflection point is a fountain of alarming risks. Digital platforms in particular are vulnerable to three major AI risks, as those in control of global platforms cede control to AI, of the algori...
arXiv Machine Learning
Jul 1
Efficient Public Verification of Private ML via Regularization
arXiv:2512. 04008v2 Announce Type: replace Abstract: Training with differential privacy (DP) guarantees dataset members that they cannot be identified by users of the released model.
arXiv Machine Learning
Oct 8
Efficient Provably Private Classification with a Tabular Foundation Model
Efficient Provably Private Classification with a Tabular Foundation Model introduces PrivTab, a tabular foundation model that embeds differential privacy directly into its architecture. Pretrained on simulated datasets, PrivTab uses in-context learning to convert sensitive rows into compact, provably private summaries, enabling a single forward pass for classification. The model outperforms private linear and neural‑network baselines under moderate-to-strong privacy, shows negligible membership leakage, maintains well‑calibrated predictions, and reduces dataset fitting time by 10,000×.