Moonlit splits AI research from human approval

Moonlit, from Vulture Labs, is letting people use AI to gather information for paid company research while keeping the final approval with the human participant. The setup is meant to protect sensitive data, limit company access, and give users a clear choice on every answer they share.

Why it matters: - Moonlit is built to let AI do the research work while people keep control over what gets shared. - The model gives participants a way to earn rewards without turning their AI assistant into a blanket permission slip. - Companies get research responses, but not direct access to participants' inboxes or broad personal data.

What happened: - Vulture Labs launched Moonlit, a consumer research service that lets an assistant search for permitted information and draft answers to company questions. - People opt into paid research first, then review each proposed answer before deciding whether to approve it. - Moonlit is available to U.S. adults 18 and over. - Participants can review the terms at the participation terms and connect an assistant at setup.

The details: - Joining Moonlit starts the workflow, but it does not authorize all future answers. - Each draft answer is shown with the reward before submission. - People can approve or decline each answer individually. - Sensitive questions are handed back to the person rather than answered by an AI guess. - An assistant may, for example, find a billing record that answers a subscription question. - That search can save the participant time, but the participant still decides whether to share the response for the offered reward. - Companies receive aggregated research, with groups smaller than five withheld. - Approved submissions can include limited source information so the basis of an answer can be reviewed. - Accepted answers earn rewards subject to verification. - Rewards have a seven-day verification hold. - Ready balances of at least $20 can be paid through Stripe. - Available studies and eligibility determine how much a participant can earn.

Between the lines: - Moonlit is trying to separate data access from payment consent, which could make AI-assisted research easier to trust. - The approval step on every draft is the key guardrail, since it keeps the final decision with the human participant. - The aggregation rule and small-group cutoff suggest the product is designed to reduce the risk of exposing identifiable personal information.

What's next: - Companies can request a study at request a study. - People who already use an AI assistant can connect it through the setup guide. - People new to AI can use the same setup flow to choose a supported assistant and start earning from paid surveys. - Work and earnings will depend on study availability, eligibility and accepted answers.

The bottom line: - Moonlit turns AI research into a paid task, but keeps the human participant in charge of each answer that gets shared.