Re-run the hypothetical, but this time without the LLM labels. Because LLM disclosure will always be voluntary.

Im assuming the policy will never prevent LLM code. The policy is purely two things: self select people into these groups and determine how the people in each group will use AI.