AI Transparency and Participant Notice

Fred may use AI-assisted components to help customers organize, summarize, cluster, or review research evidence. Depending on the workflow, that can include summaries, thematic groupings, evidence links, tags, behavioral-signal overlays, or other research-support outputs.

These features are intended to help a human team move faster and inspect evidence more coherently. They are not intended to displace human review or to produce definitive judgments about a person's inner state or legal status.

AI-assisted workflows may rely on customer-provided prompts, transcripts, survey answers, comments, recordings, interaction events, or other research material that the customer has chosen to process through Fred.

Some workflows may also rely on behavioral or attention-related signals derived from recordings or interaction data where that feature is actually enabled for the relevant product path.

Where a person is recorded, analyzed, or otherwise exposed to a workflow that uses AI-assisted research analysis, the customer should provide clear notice that the session or material may be reviewed with AI-assisted tools and that related outputs may be linked into reports or evidence repositories.

Where Fred provides native notice or acceptance steps inside the product, customers should use them. Where the collection or upload occurs outside Fred's native flow, the customer remains responsible for the participant-facing notice and any required consent or other lawful basis.

Fred expects customers to keep a human reviewer in the loop when using AI-assisted research outputs. Customers should inspect source evidence, check material context, and avoid presenting AI-assisted output as a substitute for contextual judgment.

Fred's AI-assisted outputs should not be represented as definitive statements about emotion, honesty, mental state, diagnosis, identity, or protected characteristics.

Fred does not support using AI-assisted or behavioral-analysis outputs as the sole basis for high-impact decisions about employment, education, insurance, credit, housing, policing, or similar access decisions.

Customers must not configure or request workflows that would be unlawful, manipulative, discriminatory, exploitative, or incompatible with participant rights or applicable AI and privacy law.

If a customer is unsure whether a workflow fits the supported boundary, the customer should pause the rollout and request a legal or operational review before proceeding. Fred may also pause or reject a workflow if the notice, attestation, or intended use appears unreliable or incompatible with the product boundary.

Questions about this page or a particular workflow can be sent to the contact address below.