AI companies want their models to answer health questions safely. To check that, they need people who know what a safe answer looks like. That's where nurses, doctors, pharmacists and other clinicians come in.
This guide covers what the work is, who it suits, which platforms hire for it, and the catches you should know before you apply.
What the work actually is
You won't be treating patients. You'll be reading, judging and writing. Typical tasks:
- Reviewing AI answers to health questions. Is this right? Is anything dangerous, missing or out of date? Would a patient misread it?
- Writing hard questions with model answers. For example, a realistic clinical scenario and the answer a careful nurse would give.
- Writing and applying rubrics. A checklist of what a good answer must include, used to grade AI responses.
- Comparing two answers and explaining, in clear sentences, which is safer and why.
The work is judged on your written reasoning. A short, precise explanation that cites the right guideline beats a long one that hedges.
Who it suits
- Registered nurses and nurse practitioners, especially with specialty experience (ICU, emergency, oncology, paediatrics, midwifery).
- Doctors and medical residents.
- Pharmacists, who are often good at spotting dosing and interaction errors.
- Allied health professionals: physiotherapists, dietitians, lab scientists, paramedics and similar.
- Medical researchers and public health people with strong reading of evidence.
Our research found that by 2026, expert feedback from physicians, lawyers and senior engineers had become the main kind of AI training work. That trend favours clinicians.
If you're a healthcare assistant or student without a licence, there's still work. It's usually general rating rather than clinical review. See AI training jobs with no degree.
Which platforms hire for it
- micro1 lists healthcare and medical among its fields, with an AI interview by its automated interviewer, Zara.
- Ethos recruits experienced professionals, including senior people in medicine, and screens with an AI-run voice interview.
- Mercor runs specialist review roles across many fields, including health.
- Handshake AI suits grad students and PhDs in health sciences, but most of its projects must be done from inside the US.
- Turing hires domain experts as well as coders, though healthcare isn't its main strength.
Which one is hiring for your specialty changes week to week. Check the tracker rather than trusting an old listing.
What it pays
Each platform shows pay on the listing. We only quote pay we can tie to a platform's own page. Handshake AI's own page shows roles from $40 an hour up to $125 an hour, depending on the role and qualifications. Clinical expert roles elsewhere generally pay more than general rating work, but rates vary by role, so read each listing.
Pay is usually hourly, as a contractor. You're not an employee, so there's no sick pay or paid leave. Read how AI training pay works for invoicing, payouts and tax.
How to apply well
- Put your licence and specialty on your profile, exactly. "RN, 6 years ICU, adult critical care" matches better than "nurse".
- Name your country of practice. Guidelines differ between countries. Projects often want someone who knows a specific system.
- Prepare for a clinical AI interview. Expect questions about your experience and how you reason through a case. Explain your thinking out loud, as you would to a new graduate.
- Take test tasks seriously. Cite the guideline or evidence you're relying on. Flag uncertainty instead of guessing.
The catches
- Hours aren't guaranteed. Passing the screening puts you in a pool. Projects come and go. If you're waiting, read passed the interview but no projects.
- It doesn't replace clinical income. Treat it as extra work until you've had steady hours for a few months.
- You may see distressing content. Health projects can include self-harm, abuse or graphic injuries. Mercor publishes wellness resources for experts working with sensitive content, which tells you this is a real part of the job.
- Monitoring. Some platforms track time with software that takes screenshots while you work.
- ID and credential checks. Expect a government ID check and sometimes a background check of your education and work history.
Your professional duties still apply
This is the part most guides skip.
- Check your employment contract. Some hospitals and health systems restrict outside work, or need you to declare it.
- Never use real patient information. Don't put any details from your clinical work into a task, even "anonymised". Write realistic scenarios from scratch.
- Check with your regulator if unsure. AI training work isn't clinical practice, but rules on outside work and confidentiality differ by country and profession. We can't tell you what yours says.
Where to go next
Clinical expert roles go quickly when they open. See which are live now on the job list, and check the tracker to see which platforms are adding roles this week.
Some Apply links are referral links: we get a bonus if you're hired, it costs you nothing and the application is the same.