Bookmark · 9 January 2025, 08:18 It’s remarkably easy to inject new medical misinformation into LLMs - Ars Technica ml-detection