Legal
AI Disclosure
Last updated 13 July 2026
Talenthic uses AI in a few specific places. Here is exactly where, what it is allowed to do, and what stays human.
1. Where AI runs
- CV tailoring & ATS scoring — rewrites phrasing and ordering for a specific opportunity, and scores keyword coverage.
- Auto Apply packs — drafts a cover letter and answers to screening questions from your existing CV facts.
- Opportunity understanding — normalises scraped listings into structured fields and short summaries. Original source text is always preserved and shown.
- Matching — semantic similarity between your profile and opportunities.
- Social content — captions for Talenthic's own posts. No member data.
2. Provider
All inference runs on Groq (Llama models). No other AI vendor receives your data. Your content is not used to train third-party models, and we cache results so the same text is not sent repeatedly. See Subprocessors.
3. Guardrails — what AI may not do
- It cannot invent facts. Every AI-tailored CV is checked against your master CV; new employers, dates, degrees, certifications or metrics that do not exist in your source data are rejected before you see them.
- It cannot submit anything on your behalf without your review, by default.
- It never decides eligibility, rejects you, or scores you as a candidate.
4. Human oversight
Auto Apply is review-first: we prepare the pack, you approve and submit. You can inspect every AI change against the original before sending, and edit anything.
5. Accuracy
AI output can still be wrong or awkward. You are responsible for what you submit to an employer or funder. Always read the pack before sending it.
6. Opting out
Turn off AI assistance in Settings → Privacy. Search, filters, alerts, saved items and manual applications all work without AI. Matching then falls back to deterministic rules on your stated skills and preferences.
7. Questions
Email inknamestudio@gmail.com.
