When the Assistant Slowly Becomes the Decision-Maker
Why do we keep handing more decisions to AI?
AI usually enters our lives as an assistant.
We ask it to summarize an article, improve an email, organize information, or suggest a restaurant. But over time, something subtle can happen: we begin asking it to make increasingly important decisions for us.
This is called delegation creep—the gradual expansion of responsibility we hand over to automated systems.
It might start like this:
“Improve this email.”
Then:
“Tell me how I should respond.”
Eventually:
“Should I accept this job?”
“Which investment should I make?”
“Should I end this relationship?”
There is rarely a moment when we consciously decide, From now on, AI will make my decisions. Instead, the boundary between assistance and judgment slowly moves.
Why does this happen?
Because delegation feels good. AI reduces effort, decision fatigue, and uncertainty. And when its previous recommendations seem useful, our trust grows. This can overlap with automation bias—our tendency to place excessive confidence in automated recommendations.
The danger isn’t simply that AI might be wrong. The deeper risk is that we gradually stop exercising our own judgment.
A useful rule is:
Delegate the work, not the responsibility.
Let AI gather information, compare alternatives, challenge your assumptions, and explain consequences. But for important decisions involving your values, money, career, relationships, or ethics, keep yourself in the loop.
The best AI relationship may not be “Tell me what to do.”
It may be:
“Help me think better so I can decide.”
