JavaScript

Security researchers have proven that open-weight AI models can be maliciously backdoored using minimal resources and training data. This discovery highlights a critical lack of observability and verification capabilities in the current AI supply chain.

The researchers present Program-as-Weights (PAW), a paradigm that shifts LLM usage from expensive per-input API calls to a 'compile-once, run-locally' model using neural adapters. This approach allows small, specialized models to outperform massive foundation models on specific fuzzy tasks while maintaining high efficiency and privacy.

The researchers present Program-as-Weights (PAW), a paradigm that shifts LLM usage from expensive per-input API calls to a 'compile-once, run-locally' model using neural adapters. This approach allows small, specialized models to outperform massive foundation models on specific fuzzy tasks while maintaining high efficiency and privacy.

A new side-channel attack named FROST allows websites to bypass browser sandboxing by analyzing SSD latency. By using deep learning to process I/O traces, attackers can fingerprint the applications and websites running on a user's device.