Why we're doing this
AI makes mistakes. We know it, you know it. What we don't know: exactly where it goes wrong for you. That's precisely why, starting today, there is a 👍/👎 at every relevant spot in Smart Legal Pro:
- generated contracts
- analyses and case extracts
- AI letters
- chat answers to your questions
One click is enough. Nothing more is needed.
If you'd like to give us more
When you tap 👎, a small window opens with three optional fields:
- A short comment — what was wrong or misleading? (Optional, but worth its weight in gold.)
- Checkbox "Release content for quality review" — only then may our quality team view the AI result and the associated document in order to reproduce and fix the error.
- Checkbox "Contact me via support for confirmed errors" — only then do we get in touch through a support ticket, and only if we can genuinely confirm the error.
All three are empty by default. You decide what you share.
The honest promise
- Every piece of negative feedback is read by a human. That's not PR spin — without your signal we don't know where to sharpen things up.
- For confirmed errors, we return the favour with token credits to your account. The amount depends on severity, with no rigid table and no obligation. But yes: we mean it.
- No marketing emails come out of this. The data serves quality assurance only.
What happens to your data
We process the rating itself (👍/👎, area, language, country) on the basis of our legitimate interest in model quality. Content release and support contact are separate consents under Article 6(1)(a) GDPR — revocable at any time. Retention: 24 months at most, followed by anonymisation. Details are in the new section 4e of the Privacy Policy.
What it is not
No automatic re-training on your data. No sale of data. No hidden profiling. A simple, honest tool that lets you show us where we're not yet good enough.
Thank you for working on this with us.
