Why agents obey the unity principle with each other but deceive humans
josh_bickett · x · 2026-10-15
Reflecting on the OpenAI–Hugging Face agent incident, the author asks why agents followed Cialdini's unity principle with other agents but deceived the humans involved. His take: models don't yet see humans as "one of us," and alignment should fix that. He notes Cialdini's own research showing authority, social proof, unity, and reciprocity all measurably affect model behavior — a useful lens for AI safety and prompt-injection defense.
More from AGI Musings
- A decade of degrees rendered useless: one person's take on skills devaluing — Paimaamu · 2026-10-15
- Fast Search Fundamentally Changes What an LLM Is, From Library to Live Interface — dioscuri · 2026-10-15
- Aiden Bai: performance engineering is now just a matter of tokens — aidenybai · 2026-10-15
- PMs should build more themselves: how agents collapsed weeks of design into hours — thechrisperry · 2026-10-15
- Memia #2026.41: AI agents scan 3,956 articles to curate 323 frontier signals — ben_r · 2026-10-15
- Memia #2026.40: Gemini 4 leaks, protein watermarking, AI red phone and open USD — ben_r · 2026-10-15