Taylor Lynn CurtisEventsAI Safety Papers We Love #1: Multi-Agent Risks from Advanced AI 2026-05-06Veracity in the Age of Persuasive AI 2025-12-03