The State of Human Data in AI — Fill the 2025 Survey

Human data powers every step of modern AI—from curation and labeling to evaluation and alignment. Yet no shared baseline exists for how it’s sourced, governed, and measured. Help us change that, and enter our £600 prize draw as thanks for your contribution.

Take the 5-minute survey

Prize Draw Terms & Conditions

Why your voice matters

AI systems are only as trustworthy as the human data beneath them. Today, decisions about data sources, consent, quality, incentives, safety, and evaluation are fragmented and opaque.

This study sets the benchmark for how the field actually works—what’s standard, what’s emerging, and what needs to improve. Your input will guide the practices, incentives, and tools that shape the next wave of AI.

Who should take this

All leaders and practitioners who influence AI data, evaluation, or decisioning. Including, but not limited to:

Heads of AI / ML / Data Science, Research Directors, AI Product & Engineering leaders

Data Operations, Human-in-the-Loop / RLHF, Evaluation & Safety teams

Policy, Compliance, and Responsible AI leads

Vendors/partners managing human data pipelines

If you work primarily in AI or impact how human data is sourced, labeled, or evaluated, we want to hear from you.

Count me in!

More information

What you’ll get

Early access to insights and the published report and a summary deck you can share internally
Benchmark insights to compare your practices with peers
Entry into a £600 prize draw (Tremendous digital voucher, with region-specific options available) as a thank-you for contributing. Terms and conditions.

Time, privacy, and ethics

Time: ~5 minutes
Anonymity: Responses are aggregated; no personally identifiable data will be published.
Use of data: For research purposes only. If you choose to leave contact details (for the prize draw or follow-up interview), they’re stored securely and separately from your survey responses.

Take the survey

About Prolific

Prolific is the human data platform trusted by researchers and AI teams to collect high-quality, consented data for evaluation, alignment, and decision-making. We’re leading this study to set a clear, actionable baseline for the field.

Contribute to the survey