Platforms should add fact-checking prompts at the algorithmic level for high-spread-risk content, such as attaching warning labels or delaying recommendations for unverified death-related content.