RESPONSIBLE AI
Agents prepare the work. People make the decisions.
Our rules for AI at Mondeed. What the agents are allowed to do, what they're never allowed to do, and how you stay in charge.
- 01
Humans decide
No agent can accept an offer, reveal an identity, message anyone or submit a timesheet without the approval of the person it acts for. Hiring decisions are always made by people.
- 02
No badge without a person
AI skill interviews are scored against a published rubric, and a person on our team reviews every result before a badge is issued.
- 03
Explain every suggestion
When an agent suggests or removes a gig, it says why in plain language. Conflict ratings always come with reasons.
- 04
Audit trail
Every agent action that affects a match is logged, and you can see the log for your own account.
- 05
Minimum data, private by default
Agents only see the data they need for their job. Identity fields, contract notes and verification reports are kept separate from the anonymous profile companies see.
- 06
Fair matching
We test ranking for unfair patterns, exclude protected characteristics from matching, and let you see and correct the information agents use about you.
Where the AI stops
- The conflict check is guidance, not legal advice.
- Agents never negotiate or agree a rate for you.
- Agents never unmask anyone before a mutual reveal.
- We don't use your private data to train models shared with other customers.
Questions about how we use AI? Email info@mondeed.com.