Measuring how AI is reshaping human thinking, decision‑making, trust, and lived experience.
The Human Clarity Institute (HCI) is an independent research institution publishing population-scale behavioural data and longitudinal research on the human side of AI — making changes in how people think, decide, trust and rely on intelligent systems visible, measurable and comparable over time.
How do humans remain grounded, capable, and able to flourish alongside increasingly intelligent systems?
Measurable shifts in human behaviour are becoming visible across HCI’s datasets
of participants say it is harder to know what is real online. The Digital Trust 2025 dataset finds that judging what is real, credible or trustworthy has become significantly more difficult as AI-generated content increases.
Digital Trust 2025 datasetParticipants report turning to AI as a starting point when decisions feel difficult or uncertain. Decision-Making & Digital Systems 2026
Many participants feel busy and mentally engaged without making meaningful progress. Focus & Distraction 2025
Participants report growing concern about becoming reliant on AI systems over time. AI Cognitive Load, Fatigue & Decision Offloading 2025
HCI AI Identity & Behaviour Assessment
How do your AI habits compare with people like you?
Place your responses inside HCI’s research data. See how your AI behaviours compare with all participants, people your age and people who use AI as often as you do.
See how your AI habits compare with all participants, people your age and people with similar AI use.
39 questions across 9 dimensions
Where HCI’s research looks
Attention & Cognition
How focus, distraction, mental load and cognitive clarity change as AI becomes part of daily life — and what this means for attention, intentionality and everyday thinking.
Trust & Information Judgement
How people assess what is real, credible and meaningful in AI-mediated environments — and how this shapes trust, judgement and confidence.
Agency, Identity & Values
How autonomy, self-trust, identity and alignment with values and meaning shift as intelligent systems increasingly influence decisions.
Structured evidence for people and organisations navigating the human side of AI
Professionals & knowledge workers
People who rely on clear thinking and judgement and want to understand how AI is influencing their work.
AI product & research teams
Teams that need measurable evidence of trust, reliance, cognitive load and human agency.
Researchers & institutions
Organisations seeking structured, population-scale evidence across datasets, groups and time.
Individuals experiencing change
People noticing shifts in how they think, focus, trust information or use AI to make decisions.
The human side of AI cannot be inferred from the systems alone
AI increasingly shapes the conditions in which people think, decide, work and relate to the world. But changes in judgement, trust, attention and agency are not directly visible — either to the systems people use or to those who design them. They can only be seen by measuring people, consistently, over time.
HCI is built as a research system of datasets and analysis for exactly that: behavioural measurement, longitudinal observation and human-centred interpretation, using consistent methods across datasets, populations and time.
Longitudinal measurement
Tracking how human behaviour changes over time using consistent, repeatable measurement across datasets.
Population-scale data
Collected across multiple countries and demographic groups.
Consistent frameworks
Enabling comparison across datasets and over time.
Published evidence
Designed for transparency, citation and public understanding.
Published research designed to be read, questioned and cited
Data Library
Research datasets and structured summaries tracking changes in human behaviour over time.
Explore dataInsights
Research-led analysis connecting emerging changes in human behaviour with findings across HCI datasets.
Read insightsBenchmarks
Behavioural benchmarks showing how reliance, trust, verification and agency vary across people.
View benchmarks
A human reference layer for the AI era
HCI Benchmarks show how core AI behaviours — reliance, trust, verification, decision delegation and human agency — vary across people. Each benchmark is built from structured participant data and designed to be cited, compared and reused: a stable reference for what typical, uncommon and rare AI behaviour actually looks like.
They give researchers, organisations and AI teams a common baseline for the human side of AI — and give any individual a way to see where they sit within it.
Machine-ready human experience data for AI
Human Signal Lab produces structured human-experience measurement for AI evaluation, product development, workforce research, governance and alignment — capturing forms of human response that cannot be inferred from system behaviour alone.
Explore Human Signal Lab