Nearly half of adults surveyed have allowed AI to make difficult decisions on their behalf, creating a phenomenon called “decisionrot” where users lose confidence in their own judgment. Experts and users alike suggest shifting from asking AI what to do toward using it as a decision coach that encourages reflection and exploration of priorities.
- 46% of adults let AI make difficult decisions for them
- Decisionrot describes declining personal judgment from AI reliance
- Framing ChatGPT as a coach fosters informed, autonomous choices
What happened
A survey conducted by the AI micro-learning app Headway found that 46% of respondents have allowed AI to make difficult decisions for them. Additionally, 35% turn to AI during stressful moments and 21% cannot go a full week without AI-assisted decision support. The survey highlights a growing tendency to outsource decision-making to artificial intelligence, which some experts warn could weaken human judgment.
Writer and experimenter tested ChatGPT’s usefulness by first asking it directly what to do about enrolling in an evening drawing class. The AI gave a seemingly supportive but very generic recommendation. When prompted to act as a decision coach instead of making the choice, ChatGPT engaged the user by asking clarifying questions to uncover priorities, challenge assumptions, and explore trade-offs. This approach revealed how using AI interactively preserves personal decision authority.
Why it matters
Relying on AI to make decisions risks creating “decisionrot,” a state in which users stop trusting their own judgment and avoid discomfort associated with tough choices. According to experts, this dynamic can stunt the development of critical thinking and personal growth that comes from facing uncertainty directly.
The survey also found gender differences, with men reportedly more likely to use AI to bypass difficult decisions. Developing the skill to manage unpredictability and actively choose one's direction is key to mental resilience and autonomy. Using AI as a facilitator rather than a decider supports this goal, helping users gather information and reflect without surrendering control.
What to watch next
The emerging practice of engaging AI as a decision coach rather than an answer machine may influence how individuals, educators, and organizations incorporate AI tools in everyday life. Future research will likely explore methods for optimizing AI prompts that maximize user self-efficacy and minimize overreliance.
As AI adoption grows, monitoring its effects on cognitive skills and decision-making confidence will be essential. Developers might also introduce features designed to prompt deeper user involvement and tailored questioning to help users gain insight instead of quick fixes. Users will benefit most when technology empowers rather than replaces their critical thinking processes.