AI Governance
IQNexus+ applies AI to accelerate hiring, not replace human judgment. This page describes how we deploy, monitor, and govern AI features.
AI usage policy
- AI assists candidates (resume help, interview prep, career planning) and recruiters (search, ranking, drafting).
- AI never makes a final hiring decision. It produces suggestions that humans review and act on.
- AI features are opt-in at the user level; usage consumes AI credits and is logged in your account.
Human oversight
- Recruiters and employers must review and confirm AI-generated rankings, screening notes, and outreach before sending.
- Candidates can always disable AI-assisted ranking on their profile.
- Admins audit AI feature usage via the Admin → AI Credits dashboard.
Bias mitigation
- Models are evaluated against diverse evaluation sets before promotion.
- Talent scoring excludes protected attributes (name origin, photo, age, gender, ethnicity).
- Recruiters see ranking factors, not opaque scores; rankings can be challenged and corrected.
- We monitor outcome distributions across cohorts to detect drift.
Data handling for AI
- Candidate data is not used to train third-party foundation models.
- Prompts and completions may be retained for up to 30 days for abuse detection and quality, then deleted.
- Sensitive identifiers (email, phone) are redacted from prompts where possible.
- Inference runs through the Lovable AI Gateway; underlying providers are listed in our sub-processor list.
Model transparency
- Every AI feature labels which model family was used.
- Users can see what input was sent (prompt preview) before AI runs.
- Model changes are announced in release notes; major upgrades may pause and require re-consent for sensitive flows.