Transparency

AI Disclosure

Last updated: 7/21/2026

AiSA is a directory of AI implementation specialists — so it's only fair that we're clear about where we use AI on our own site, what model powers it, and how humans stay in the loop.

Where AI is used

• Provider onboarding. A conversational assistant helps providers structure their profile and drafts a polished bio paragraph. The provider reviews and edits before publishing.
• Homepage AI search. A chat interface answers natural-language queries about which providers might be a good fit, based on published profile data.
• Per-profile Q&A. On each provider profile, a scoped chat answers questions strictly from that provider's public profile data.
• Recommendation drafting. After a completed engagement, a short chat with the buyer produces a draft "Recommended by" write-up. The buyer reviews, edits, and approves before anything is published.

Which model

We currently use Google Gemini models via the Lovable AI Gateway. The model may change over time as better options become available.

Human review

All AI-generated content that appears on a person's profile — bios and recommendation entries — is reviewed and editable by the human it represents before publication. AI search answers and profile Q&A are generated on demand and may contain errors; treat them as a starting point, not authoritative advice.

Data handling for AI features

AI conversations are logged in anonymized form for rate limiting and product improvement. We do not use your conversations to train third-party foundation models. See our Privacy Policy for details.

Limitations

LLM outputs can be confidently wrong. Buyers should always verify claims directly with a provider before entering any engagement, and providers should double-check any AI-drafted content that appears on their profile.