Does having used AI before change whether you feel more in control, or more unsettled, when it shows you how it works?

This explores whether past experience with AI decides who feels more in control when an AI shows its inner workings. The short answer is that the corpus doesn't test prior AI experience as a predictor directly. What it does show is that the same act of transparency splits people. When researchers showed users an AI companion's system prompts and conversational strategies, some users felt empowered and others felt manipulated. The split came from individual differences, not from the disclosure itself Does revealing an AI's strategies help or harm user agency?. So the premise of the question holds up. Which trait does the sorting is still an open question here.

The nearest evidence is about experience gained during an interaction, not before it. When people learn they're working with an AI, they first steer away from it. That bias reverses only after repeated rounds where they can see the results. Telling people it's an AI, with no feedback on results, calibrates nothing Does revealing AI identity help or hurt user trust?. This suggests a reframe: what counts may be seeing results that link what the AI says to what it actually does, not how much AI someone has used before. Transparency may empower people only once they have a track record to check it against.

A second idea from the corpus is that feeling in control and being in control are different things. Users of agentic AI trust it when they can visibly step in, even if they rarely do Does perceived control matter more than actual control for AI trust?. People feel they own AI-written text when they shaped it, while personalizing the model makes no difference Does user control over AI text shape feelings of ownership?. Applied to transparency, disclosure may feel empowering when it hands people something they can act on, and threatening when it only reveals strategies they can't push back against. Leonardi's work adds that crediting AI with its own judgment can quietly reduce people's sense of control, even when they keep formal authority Does attributing agency to AI actually reduce human sense of control?.

There's also a warning that confidence is a poor stand-in for real understanding. People felt equally sure about their privacy whether or not they had actually removed identifying details Why do people feel safe sharing when they haven't actually removed sensitive information?. People also often preferred AI replies that leaked private information Do people notice when AI agents leak private information?. Separately, people who work with LLMs can mistake the AI's output for their own skill How does AI-assisted work reshape how people see their own abilities?. If experienced users feel empowered by transparency, that feeling could reflect familiarity rather than better judgment.

What you might not expect: the more useful question may be what users can do with transparency, not who they are. The corpus points to feedback loops and the ability to step in as the conditions that turn disclosure into empowerment. A direct test of prior AI experience as a moderator is a gap in this collection.