AI With a Human Signature
We use AI to work faster. A named person checks the evidence and signs every paid deliverable, and every deliverable tells you how AI was used.
How We Use AI
We use third-party large language models from established providers to speed up research and drafting. Each deliverable names the models used in its AI disclosure. The providers are listed on our Sub-processors page.
AI helps us find and organise evidence. People make the calls: a named verifier checks the evidence, decides the verdict and signs every paid deliverable. Every decision-critical and high-impact claim is personally re-checked by the verifier before signing.
Signed and Unsigned Outputs
- Signed by a named verifier: Build Verdict, Release Assurance Pack, Release Assurance Suite, Decision Brief and Go-to-Market Build.
- AI-generated and unsigned: the free Offer Signal, First Signal and AI Visibility Signal, and the Validation Brief. Each is labelled as AI-generated and is not for investor or board use.
If a paid deliverable reaches you without a named signature, it is not a finished ThriveFinity deliverable — tell us.
What Every Deliverable Discloses
- the AI models used;
- the dates the research was carried out;
- which steps used AI and which were done by a person;
- the name of the verifier who checked and signed it (paid deliverables).
EU AI Act Article 50 — disclosure-aligned, not a certification
Article 50 of the EU AI Act sets transparency duties for AI-generated content. We use models built by others, and our paid deliverables are private reports to a named client, so most of Article 50 will not apply to a typical deliverable. We disclose anyway, so the disclosure travels with any part a client publishes. “Disclosure-aligned” is not a certification or conformity assessment, and makes no claim about other AI Act obligations. The UK has no equivalent statutory duty today; we apply the same disclosure everywhere.
What AI Does Not Do Here
- Sign or approve a paid deliverable
- Give regulated financial, investment, legal or medical advice
- Use unattributed sources in a signed deliverable
- Train on your documents or briefs
- Replace a qualified lawyer, accountant or financial adviser
Your Data
We do not use your material to train AI models. Project material is kept for 90 days after delivery and then deleted, unless you ask us to delete it sooner. Records we must keep by law, such as invoices, are kept for the period required by applicable tax and company law.
To access, correct or delete your data, see Your Rights in our Privacy Policy or use the Privacy Centre. We acknowledge requests within 2 working days and respond within one month.
Raise a Concern About AI Use
If you have a question or concern about how AI was used in your deliverable, email om@thrivefinity.uk with AI concern in the subject line. We acknowledge within 2 working days and respond within one month.
Questions About Our Use of AI?
Ask us. A named person reads and answers every question.