Read the rule before opening a tool
Research may examine your interpretation, recall, language or decisions. An AI tool can change the response the researcher receives even if its wording sounds plausible. Do not assume that a tool you use for other writing is permitted in a paid activity, or that an invitation mentioning AI authorizes you to generate answers.
Prolific's participant guidance prohibits AI assistance unless explicitly requested by the researcher. That is its platform rule. Check the rules of your actual provider and the specific study before starting. If they are unclear, ask rather than treating silence as permission.
Preserve the experience being studied
When a prompt asks why you chose a product, describe your actual reasons. A generated explanation may sound coherent while adding motives you never had. Rewriting can also remove hesitation, uncertainty or distinctive language that the research is trying to understand. More polished prose is not automatically more accurate evidence.
Hypothetical example: you found a label confusing but cannot immediately explain why. Writing a brief honest uncertainty is different from submitting a tool's detailed criticism of colours and layout you did not notice. The concept testing guide explains why interpretation should remain tied to the information and experience you actually had.
Keep screeners and profile facts your own
Do not use AI to invent qualifying work experience, predict the answers a recruiter wants or generate several identities. A screener asks whether your circumstances fit the study; it is not a writing competition. Rephrasing inaccurate information does not make it a truthful application.
UserTesting's contributor code calls for honest screeners and participants' own responses, including restrictions on generated responses. Its rules apply to its contributor process. Use honest screening guidance to describe your role, use of products and availability precisely. Leave an unsuitable study rather than constructing a background that would qualify.
Check the exact scope of permission
A study can deliberately ask you to interact with an AI system. Read which system, which steps and which outputs it wants. Permission to test a supplied chatbot is not automatically permission to use another model to answer the final questionnaire. Follow any instructions about reporting tool use or preserving your initial response.
If the researcher authorizes assistance, keep within that scope. Ask before changing tools, adding an extension or submitting generated text where your own reflection was requested. A useful clarification states the action you want to take and the instruction you are uncertain about. Do not send confidential prompts to an outside tool merely to formulate that question.
Treat accessibility and privacy explicitly
If you need translation, dictation, reading support or another accommodation, ask the provider how to use it within the study's requirements. Tools differ, and a function labelled writing help may generate new content. The provider should explain the permitted arrangement for the particular task rather than leaving you to guess.
Also check whether a tool sends text, recordings or screenshots to another service. Study material can be confidential even when it contains no obvious personal identifier. The research privacy checklist helps identify boundaries. This guide does not decide which accommodation a provider must offer; it encourages clarification before the activity begins.
Respond accurately if something went wrong
If you used assistance before noticing a restriction, stop and contact the provider through its official route. Explain which tool and stage were involved without inventing a justification or resubmitting repeatedly. Ask how the existing response should be handled. The provider decides the consequence under its rules; a new answer does not automatically erase the earlier submission.
Scoutela provides free public previews and paid provider details and application access under current pricing. It does not authorize AI use in another organization's study or override its review. Choose tasks you can complete within their permitted boundaries, and keep access to a discovery service separate from permission to alter the research response.
Common questions
Can AI make my screener answers more likely to qualify?
Do not generate or alter facts to qualify. Give accurate answers yourself and follow the provider's assistance rules.
Is AI allowed when the study is about AI?
Only within the activity's explicit instructions. The topic alone does not authorize assistance with every stage.
What if I need language or accessibility support?
Ask the provider before starting which assistance is permitted and how it should be used or disclosed.
Sources and further reading
Provider-specific policies apply to that provider. Check its current terms before taking part.
- AI assistance in submissions Prolific
- Contributor Code of Conduct UserTesting
