Image: Anthropic
Anthropic wants to hear how you really feel about AI, and this time your answers don't have to stay private. Its Claude-powered research tool, Anthropic Interviewer, is inviting Claude users to a roughly 15-minute conversation. When it ends, you can choose to make your whole interview public for anyone to read.
TestingCatalog reports that Anthropic has started prompting more users to take part. Its screenshots from the Claude app show a card titled "Use your voice to help guide AI" with a "Begin the interview" button. TestingCatalog calls it a voice interview, but the card itself says "chatting," and Anthropic's study page doesn't mention voice at all, so "voice" may just be a figure of speech.
Anthropic started prompting more users to participate into a 15 minutes voice interview via its Anthropic Interviewer.
— 🚨 AI News | TestingCatalog (@testingcatalog) October 4, 2026
> “Excited, worried, or somewhere in between? We’re interested to hear how AI fits into your life and your hopes for what comes next.”
Later, users may… pic.twitter.com/ymJacn9jW7
What Anthropic has confirmed
Anthropic announced the study on its own "What do you want from AI?" page. It's open to Free, Pro and Max users on Claude and Claude Code whose accounts are at least two weeks old, and it closes on October 6. The interviewer asks about your most meaningful experiences with AI, good and bad, what you'd like AI to help change in areas like work, school or healthcare, and what you want from AI companies, Anthropic included.
Going public is optional. If you opt in, Anthropic publishes the complete interview, without edits, along with your country, but not your name or email. It warns that small details like your job or city could still identify you, and that you should treat the choice as permanent, since it can't delete copies other people have saved. You can end the interview early, and the partial record is deleted.
From 1,250 interviews to 81,000
Image: Anthropic
The tool isn't brand new. Anthropic introduced Anthropic Interviewer with a test of 1,250 professionals. Among the 1,000 general workers in that test, 86% said AI saves them time, yet 55% voiced anxiety about what AI means for their future.
A week-long pilot followed, and 80,508 people across 159 countries and 70 languages took part. Back then, only high-level results and a few quotes were made public. Anthropic also runs a monthly Economic Index Survey through the same tool, asking a small, random group of users how AI is changing their work.
Why the public option matters
Making interviews public turns a company survey into a shared record. Anyone will be able to check whether Anthropic's findings match what people actually said, which is a real check on the company running the study. Anthropic says nothing like this has been made public before.
The fine print is worth a look before you start. Anthropic's privacy note says it may combine your answers with privacy-preserving summaries of your Claude usage. If you use Claude's Memory feature, your answers may also be added to it, depending on the choices you make during the session. Even private interviews still feed Anthropic's research. Anthropic also admits that everyone in the sample uses Claude, so the results won't reflect the public as a whole.
TestingCatalog thinks the interviews will soon reach a broader audience. Anthropic hasn't confirmed any expansion beyond this study, but after a test, a global pilot and a monthly survey, Anthropic Interviewer is clearly here to stay. If the card pops up for you, take it. Just keep names and personal details out if you plan to go public.
Comments