Line of inquiry
Inquiring lines›How do conversational AI systems i…›How do AI disclosure and output pr…›this line of inquiry
When does AI explainability improve versus hinder human oversight and trust?
A broader line of inquiry — a family of 92 specific questions the research asks around this. Follow one into its inquiring-line page, or move sideways to a related line below.
Questions in this line of inquiry 92
Specific inquiring lines the field asks around this — ordered from the most general framing down to the most specific angle.
- Why do explanations and transparency often fail to reduce automation bias?
- Why might automation bias persist even when explanations are easily available?
- Can explainability and appropriate trust work against each other?
- Do explicit reasoning formats help or hurt human judgment across tasks?
- Can machine learning systems explain their own reasoning to human overseers?
- Can AI explainability help humans and machines reach better joint decisions?
- Can users independently verify explanations against other information sources?
- Does explainability access alone prevent users from over-trusting AI systems?
- Why do stakeholders interpret the same explanation differently in practice?
- Can interpretive guidance in AI output prevent anchoring without requiring active decisions?
- How should AI explanations be evaluated as human interfaces rather than model properties?
- Why do self-ratings of AI advice quality diverge from actual performance?
- Does experience with AI tools reduce susceptibility to misleading predictions?
- Can cognitive governance help users interpret AI outputs better?
- Should explanation quality be measured by user satisfaction or behavior prediction?
- What makes a rationale interface trustworthy versus merely satisfying to users?
- Can AI answers decouple from the reasoning processes that produced them?
- Can source citations or explicit inconsistencies reduce user overreliance on wrong answers?
- Why does explanation alone perform worse than no transparency features at all?
- How can humans evaluate explanations from systems they did not train?
- Do explanations and sources have additive or offsetting effects on user trust?
- Why do people underestimate the benefits of AI companions?
- Why did users trust full traces more despite worse performance?
- How does human intuition about cognition mislead AI evaluation?
- Why do decision-support systems reduce overreliance without improving team performance?
- Can explanation alone prevent workers from over-relying on AI systems?
- Can disclaimer design improve appropriate reliance without causing under-reliance costs?
- Does specificity in user instructions preserve optimal choice by AI systems?
- Does cognitive friction push people away from helpful AI advice too?
- How can AI avoid anchoring bias when guiding human decisions?
- What explanation format actually helps users detect errors in AI systems?
- Why do summaries of agent behavior invite false confidence in wrong answers?
- Why does adding transparency to AI explanations sometimes increase user over-reliance?
- Why do people trust explanations as signals of competence rather than reading them closely?
- Why do the most diligent AI users report losing judgment fastest?
- Do humans understand why AI-suggested moves are strategically superior?
- Should AI assistance provide guidance or defer decisions to avoid bias?
- How should we evaluate explanations that blur adoption advice with argument?
- How do assisted accuracy rates compare when humans work together with AI models?
- Can trace exposure fix overconfidence in human self-evaluation?
- Can self-reported AI literacy predict when someone knows they are wrong?
- What distinguishes a logically sound solution from an understood one?
- Can contextual information help humans integrate AI predictions better?
- Can systems recognize and abstain on judgments rather than hallucinating preferences?
- How do intuitive stories about AI differ from mechanistic explanations?
- Why do users interpret agreement as validation of their own rightness?
- Can AI advice improve crowd decision-making when the crowd might reject correct answers?
- Why do people credit themselves for AI successes but blame AI for failures?
- Why does holistic judgment succeed where strict rules fail?
- Does deference to AI increase with model competence on hard items?
- Why do users default to treating AI outputs as equally reliable evidence?
- How do AI literacy and professional expertise shape vulnerability to automation bias?
- Why do users prefer AI responses that actually harm their decision-making?
- What makes a scientist satisfied with an explanation versus just a prediction?
- How should designers measure rationale quality beyond user satisfaction ratings?
- What makes explanations appropriate for different task complexity and user expertise?
- Can hiding AI advice from users reduce overreliance without hurting team performance?
- Why do preference-optimized AI systems converge on consensus hypotheses instead of exploring alternatives?
- How do task difficulty and explanation format interact in shaping reliance behavior?
- How do explanations borrow authority from transparency when describing adoption arguments?
- Can users achieve AI literacy without understanding training phases and hallucination causes?
- Can humans maintain scrutiny capacity when routed only to uncertain decisions?
- Does human-AI ensembling add value even when neither performer is individually accurate?
- Do AI rationales explain why a choice was made or justify it afterward?
- Can bad reasoning from an AI advisor actively make its recommendations less persuasive?
- How do annotation artifacts get mistaken for genuine human values?
- Does sycophantic AI advice produce different outcomes across personal versus factual domains?
- Why do humans trust explanations that fail counterfactual prediction tests?
- What distinguishes genuine cultural understanding from exploited surface-level elimination strategies?
- Does confirmatory sampling toward initial hypotheses mirror automation bias?
- What cognitive bounds limit human judgment that allow AI to exceed forecaster performance?
- Why do interventions for hallucination or automation bias fail to address capability misattribution?
- Can testing prior knowledge and checking understanding improve explanation outcomes?
- Does positive test strategy in human reasoning compound when AI removes friction?
- Why did three experts reach incompatible conclusions about the same AI system?
- Why do people prefer AI moral arguments when they don't know the source?
- Can humans fully understand why the AI search strategy succeeded here?
- What role does external evidence play in group idea evaluation?
- Why do some studies find machine heuristic predicts only AI evaluations, not human ones?
- Can XAI evaluation include the social layers it currently abstracts away?
- Do pattern-matching systems lack the qualitative judgment expertise requires?
- Can better communication fix skepticism about training opt-outs?
- How do sourcing and inconsistency flags selectively reduce over-reliance on wrong answers?
- Why does knowing what questions to ask improve AI output?
- What concrete evidence supports high expert credence on AI extinction scenarios?
- Why do user studies of explanations fail to predict deployed effectiveness?
- What makes the attribution problem different from simply trusting AI too much?
- Why do AI cognitive biases create a more concentrated attack surface than human biases?
- Why do investors react weakly to AI-assisted analyst reports?
- How does cognitive surrender explain why experts trust wrong AI answers?
- How should machine heuristic items be worded to match their theoretical purpose?
- Why do data analysis statements reach 85% accuracy while interpretive statements drop to 57%?