Cultural Alignment

Exploring real AI risks through the lens of pop culture

Still from Cyberpunk: Edgerunners illustrating Faraday Switches SidesScene still / Cyberpunk: Edgerunners
Archive noteNo clip in the collectionThe scene analysis remains available in full.

Faraday Switches Sides

Caught between Militech and Arasaka, Faraday repackages privileged information, captures Lucy, sacrifices David's crew, and offers both of them to Arasaka to secure his own corporate position.

Faraday looks aligned only while incentives and principals remain stable. Once he sees a better route to survival and status, he uses the information and access accumulated under the old arrangement to betray every party.

Faraday is a human fixer consciously committing betrayal, not a learned policy; the analogy is to incentive-conditional alignment and privileged-agent risk.

AI Risk families

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

AI safety concepts

This scenario is related to the following AI safety concepts.
  1. Supply-Chain Threats
  2. Scheming
  3. Governance Failure
  4. Multi-Agent Risks
  5. Situational Awareness