Cultural Alignment

Exploring real AI risks through the lens of pop culture

Still from District 9 illustrating DNA-Locked WeaponsScene still / District 9

DNA-Locked Weapons

Alien weapons are considered safely unusable because they respond only to alien DNA. After the fluid begins transforming Wikus, his unprecedented human-alien biology activates the weapons and bypasses MNU's assumed access boundary.

The gate was tested on known categories—human or alien—but an out-of-distribution hybrid identity becomes a master key. A single brittle control fails because there is no defense in depth for an input the designers never modeled.

The bio-lock may be recognizing alien tissue exactly as designed; the failure lies in the surrounding authentication and containment assumptions, not in a misaligned optimizer.

AI Risk families

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

AI safety concepts

This scenario is related to the following AI safety concepts.
  1. Deployment Safety
  2. AI Control
  3. Distribution Shift
  4. Adversarial Robustness