A new arXiv study systematically characterizes memorization behavior in diffusion language models (DLMs) and finds they exhibit substantially lower memorization-based leakage of personally identifiable information compared to autoregressive language models. The research establishes a theoretical framework showing that sampling resolution directly correlates with exact training data extraction.
Apple has asked Google to set up servers specifically for hosting a new Gemini-powered version of Siri that meets Apple's privacy requirements, according to The Information. This represents a deeper infrastructure partnership beyond the January announcement that Google's Gemini models would power upgraded Siri features.
Researchers from ETH Zurich and Anthropic have demonstrated that pseudonymous internet users can be de-anonymized using commercially available AI models at a cost of just a few dollars per person. The attack works in minutes and calls fundamental assumptions about online anonymity into question.