NewFour certification pathways are now open. See the certifications ↗
the human union✦

Certifications

Model training & evaluation

Teach and test AI models: judge their answers, find their failures, and check their expertise.

RLHF / preference raterAvailable now

Certified in

  • Rating AI answers against a shared rubric: helpful, honest, harmless
  • Checking facts before judging style
  • Writing specific, checkable reasons for each rating
  • Flagging unsafe content instead of scoring it
Get certifiedLearn the role and earn the certificateI already am oneShow your credentials and who you've worked with

Red teamer / safety testerAvailable now

Certified in

  • Finding AI failures and sorting them by type of harm
  • Rating how serious each failure is
  • Writing reports another tester can repeat step by step
  • Staying in scope and reporting only through agreed channels
Get certifiedLearn the role and earn the certificateI already am oneShow your credentials and who you've worked with

Domain-expert reviewerComing soon

Certified in

  • Checking AI answers in their own field, such as medicine, law, finance or engineering
  • Backed by a recognised professional qualification in that field
Get certifiedJoin the list. We'll tell you when it opensI already am oneShow your credentials and who you've worked with

Every Human Union certificate has a unique ID that anyone can check at Verify a certificate.