Fetching the paper…

CI-Bench: Benchmarking Contextual Integrity of AI Assistants on Synthetic Data · Around