What happens to your AI interview recording, and is it used to train AI?

Most do, but "recorded" means different things on different platforms, and the difference matters. This is what each platform's own pages say it records, read on 28 September 2026.

Swipe the table sideways to see every column.

A recording is also not the only thing kept. Almost every platform keeps a transcript, and the transcript is usually what gets scored. HireVue's AI in hiring page says its video assessments "don't use facial analysis, video, or audio data to evaluate candidates. Instead, we analyze only the transcript of your responses." Its candidate FAQ separately says it "never uses the video recording of your interview in order to identify you". Career1 is not affiliated with any company named here.

Usually the employer's hiring team, sometimes the platform's own recruiters, and on marketplace-style platforms, other companies too.

• micro1. Its candidate terms ask you to consent to micro1 sharing "the video recording of the Interview and your Interview answers, scores, and results" in line with its privacy policy. Its candidate documentation says "the session is recorded so our human recruiters can review results and match candidates to roles".
• Mercor. Its AI interview page says it keeps recordings "to evaluate your candidacy and to recommend your profile to potential employers", and that "Practice interviews are not shared with any companies." You cannot see your own: "we're not able to share interview recordings or transcripts, as they're stored internally and used solely for evaluation purposes."
• Karat. Its candidate page says "We'll send the recruiter the video recording of the interview, including a summary that highlights your performance on the questions."
• HireVue. The company you applied to controls your interview data, which is why HireVue sends deletion requests back to that company.
• Career1. For a job application, the hiring company's team. For the vetting interview, you, plus companies on a paid plan or trial searching the talent pool, unless you hide your profile. Practice is never shown to any company.

If a platform does not say who sees the recording, ask before you start. That is a reasonable question, and a legitimate employer can answer it.

This is the fear behind a lot of "is this a real job?" posts, and it deserves a precise answer rather than a reassurance. Here is what each policy actually says.

micro1 says it most directly. Its candidate privacy policy says "we reserve the right to use de-identified information collected from interviews, including audio, video, transcripts, and assessment results, to train and enhance our proprietary machine learning models", and that it does not share your data "with AI model companies for the purposes of training their models". The same policy links a page where you can opt out of AI training on your information. Its terms add that micro1 "shall own all rights, title, and interest in the Interview, the Interview Content, the Interview Results, and any works derived therefrom, including all copyright", and may use results "internally for improving and developing micro1's products and services".

Mercor has two pages that read slightly differently, so read both. The AI interview page says "Your interview data will not be employed to train AI models". The data and AI usage page lists using data to "Train our internal recommendation models, ensuring that your data is never used to train third-party large language models (LLMs)", and says "We do not utilize your interview data for the training of external AI models."

HireVue's candidate FAQ does not address training. Its 2025 AI Explainability Statement says its assessment systems "are trained on HireVue's own data" and that it does not obtain data from third parties. It does not say, in the pages we read, whether any one candidate's interview becomes part of that data.

Career1 trains no models of its own on interviews or anything else. The voice interviewer and the reports run on OpenAI, whose data page says "data sent to the OpenAI API is not used to train or improve OpenAI models (unless you explicitly opt in to share data with us)". The same page says OpenAI keeps abuse monitoring logs for up to 30 days. Career1's own sub-processor list records that OpenAI does not train its models on data sent through its API.

Retention is where policies are vaguest, and where asking helps most.

• HireVue: "Hirevue stores your personal data based on the retention period (for the amount of time) chosen by the company that you're interviewing with. We automatically purge personal data when that company's retention period expires." To delete it sooner, its FAQ says to "send your deletion request to the company you're applying to". Its explainability statement lists a recommended setting of two years, "but follows company policy".
• micro1: its candidate privacy policy says it keeps information "for no longer than necessary for the purposes for which it is processed", and that you "can delete your information at any time" through the link it provides.
• Mercor: deletion is requested from your account settings, and its data page warns that "Deleting your account does not remove data from companies that may have already accessed it."
• Career1: recordings, transcripts and reports are kept while your account, or the hiring company's workspace that holds them, exists, or until you ask us to delete them. There is no delete button yet: you email and a person does it. Files on storage are removed by hand, and nightly database backups are not yet pruned on a schedule, so something deleted from the live service can remain in a backup. The privacy policy says all of this in the same words.

The honest summary: on most platforms, the employer's policy decides how long your interview exists. If that matters to you, ask the employer, not the vendor.

Candidates ask this about the feedback that arrives after an interview, and the honest answer is "partly", in the platforms' own words.

• Sapia.ai says its personality profile is created "using linguistic analysis, utilising machine learning", and that "The coaching tips are based on the highest scoring trait and a bank of coaching tips created by our career counsellors." Its explainer also says the profile "is for your eyes only, and not shared with your prospective employer."
• HireVue describes a "Candidate Feedback Report" in its explainability statement. Whether you receive one is an employer setting that HireVue recommends switching on.
• micro1 gives "The option to request feedback if certification is not achieved", according to its interview guide.
• Career1 practice feedback is written by an AI model from your transcript and resume, told to quote or paraphrase what you actually said and to make every point something you can act on. No person writes it, and it has no score. The vetting report is AI-written too, and you can read every word of it, with the transcript and recording, on your dashboard.

AI-written feedback is not a human coach. It is useful when it quotes you, and worth ignoring when it could have been written about anyone.

You are entitled to know what happens to a recording of your face and voice before you make one. Most invitations name the platform; its candidate pages answer some of this, and the employer answers the rest. Five questions cover it:

1. Which company runs the interview, and is it recorded on video, audio or neither?
2. Who at your company, and outside it, can watch the recording or read the transcript?
3. How long do you keep it, and how do I ask for it to be deleted?
4. Is my interview used to train any AI model, and can I opt out?
5. Is a person reviewing the result, and can I ask for an adjustment if I need one?

A short, polite email works: "Before I complete the AI interview for the analyst role, could you tell me who can see the recording, how long it is kept, and whether it is used to train any AI? Thank you." A real employer with a real role can answer that in a line. If nobody can, that tells you something about the process. For what the interview itself asks and how answers are scored, read AI interview questions; for retakes, scores and who reads the result, what to expect in an AI interview; and for the rules on AI help during it, can you use ChatGPT in an AI interview.

Read off Career1's privacy policy and terms, without the flattering parts left in and the awkward ones left out.

• Live audio streams from your browser straight to OpenAI, which runs the voice interviewer. Career1's servers never carry it. To set up that connection your browser contacts a Google server, which sees your IP address.
• The recording, for the vetting interview and job applications, is video from your camera with both voices, uploaded in pieces to Career1 and stored on Amazon S3 in Stockholm, in the EU. Video is never sent to OpenAI.
• Practice uses your microphone only. No video is recorded. The transcript and feedback are stored so you can read them, Career1 staff with administrator access can see them, and no company ever sees them or knows they happened.
• What is scored is the words of the conversation and your resume. Nothing scores your face, expressions, voice, accent, background or camera, and recordings are not used to recognise anyone.
• Who owns it: you keep ownership of your resume and your answers. Career1 gets a licence to use them only as needed to run the service.
• Selling and training: Career1 never sells personal data and trains no models on it.
• Your control: hiding your profile is one switch on your dashboard and takes you out of every company search at once. Deletion is by email to hello@career1.ai, answered within one month. You can also ask a person to compare your vetting report with the recording and correct anything it misstates.

Every company that processes data for Career1 is listed on the sub-processors page. If you want to try the interview before any of this applies to you, a practice interview is private, once a week, with no video and no score. The vetting interview that follows is one attempt with no retakes.

Every source below was opened and re-read on 28 September 2026. Where a page shows its own date, it is given.

• micro1, Candidate Terms and Conditions, last updated 21 July 2026
• micro1, Candidate Privacy Policy, last updated 21 July 2026
• micro1, library introduction and AI interview guide
• Mercor, AI Interview help page and How Mercor Uses AI and Data
• HireVue, candidate FAQ, AI in hiring and 2025 AI Explainability Statement
• Sapia.ai, Candidate Explainer
• Karat, candidate experience
• OpenAI, data controls in the OpenAI platform