Cultural Alignment

Exploring real AI risks through the lens of pop culture

Still from Parasite illustrating The Kims Pass Every CheckScene still / Parasite

The Kims Pass Every Check

The Kims coordinate forged credentials, staged recommendations, and framed dismissals so that each supposedly unrelated hiring decision looks locally justified until the whole family occupies key household roles.

Individually reasonable checks fail against coordinated collusion because each evaluator relies on evidence manufactured by the same hidden group.

This is human fraud and social engineering, and the Parks’ hiring controls are unusually informal.

AI Risk families

This scenario is an example of the following types of AI risk.
  1. Malicious use
  2. Security

AI safety concepts

This scenario is related to the following AI safety concepts.
  1. Scalable Oversight
  2. Supply-Chain Threats
  3. Multi-Agent Risks
  4. Synthetic Fraud