WHIZZSCRIBE

A Whizz Tech product · Whizz Scribe

Your competitors publish every day. You publish when there's time.

Whizz Scribe is an AI content writer that researches your topic, writes in your brand's voice, and publishes straight to your site — on its own.

I'm a developer — open the tool →

3 minutes

from idea to a published, cited article — not 3 weeks

From $29/mo

less than a single agency blog post

Real sources

researches live and cites what it claims — no made-up facts

Legal

AI Disclosure

Effective 1 July 2026 · Version 1.1

1. What the system is

Whizz Scribe generates long-form articles using large language models (currently supplied by OpenAI), combined with live web research and retrieval from a knowledge base you provide. Every article is machine-generated text, produced at your request and grounded in sources the pipeline actually found — the citation list on each article is the record of those sources.

2. Our stance on EU AI Act Article 50

Article 50 of Regulation (EU) 2024/1689 requires that AI-generated or AI-manipulated text published to inform the public be disclosed as such — unless the content has undergone human review and a person or organization holds editorial responsibility for its publication.

We built the product around that exemption path rather than against it:

  • Human review by default. Generated articles enter your workspace in an in_review state. Nothing publishes without an explicit action by your team — in the dashboard or via an authenticated API call. Editorial responsibility stays with you, the publisher.
  • Provenance metadata on every article. Each article object carries its full generation record: the pipeline stages that ran, the citations found during research, quality scores, and structured JSON-LD identifying the publishing organization. This metadata is available over the API so you can retain it in your own systems.
  • Disclosure stays your call — and your duty. Where your use does not meet the human-review and editorial-responsibility conditions, or where another law requires labeling, you must disclose AI generation to your readers. Many publishers choose to disclose regardless; the metadata we attach makes that straightforward.

3. Accuracy and its limits

The pipeline is designed to reduce — not eliminate — model error: research runs against live sources, unverifiable claims are dropped rather than invented, external claims keep inline source links, and a strict quality rubric flags weak drafts. Language models can still be wrong. Treat the review step as real editorial work, especially for medical, legal, or financial topics.

4. What we do not do

  • We do not train models on your knowledge base or your generated articles.
  • We do not auto-publish anything without an instruction from your account.
  • We do not fabricate citations — a claim without a verified source ships without a link, or not at all.
  • We do not generate content designed to impersonate real people or conceal its commercial origin.

5. Questions

If your compliance team needs specifics — subprocessor details, data flows, or the exact provenance fields on the article object — write to support@whizztech.ai. Engineers answer this inbox.