In one sentence
A good language model is now common ground — the question is no longer whether an AI answers, but how carefully. That is exactly where the difference is made: in the supposed small things that separate “sounds plausible” from “holds up under scrutiny”.
Same model, a different result
Smart Legal Pro uses the same generation of large language models you know from ChatGPT. You still get a different result — because around the model sits legal craftsmanship a bare chat doesn’t have. Six of those small things:
1. Ask back first, then answer
Laypeople often don’t know which details are decisive. A bare chat then guesses away. Our Guided mode flips this: the AI asks one follow-up at a time until the facts are clear — and you can ask back anytime (“What does this mean?”, “Why does it matter?”) or honestly answer “I don’t know”. Whatever stays unclear is flagged as an open point instead of being invented.
2. Catch contradictions instead of smoothing them over
Say “the termination arrived on the 3rd”, later “end of April” — a friendly chat nods to both. Our interview flags the contradiction and asks you to clarify before it becomes a letter. Small thing, big effect: a statement of facts with an internal contradiction holds up before no authority and no court.
3. Traceability: a statement of facts with a receipt
The interview produces a clean statement of facts — with a transcript. You see (and share with your lawyer) how the facts were gathered, including the points where you were unsure. A chat history is fleeting; here the derivation is part of the result.
4. Sources that actually exist
Language models love to invent case numbers and statutes. We keep it clean: standard answers invent no rulings, and on request we add verified primary sources — every link checked for reachability, often with an archive snapshot. If the research fails, the surcharge is refunded automatically.
5. Your data stays your data
Before any text goes to a model, we mask personal data (names, addresses, file numbers) and only reinsert it in the finished output. In an open chat window, your client data lands unprotected at the provider.
6. Law isn’t the same everywhere
DE, AT, CH — and more — differ in deadlines, statutes and wording. Through playbooks and a fixed jurisdiction, the AI gets the right basis instead of a vague “German-speaking” blend. A small thing like a wrong deadline can cost the entire case — which is why critical points also get a deadline second opinion.
Staying honest: we are AI too
So there’s no misunderstanding: Smart Legal Pro is an analysis and drafting tool, not a lawyer, and does not replace legal review. The point isn’t “AI good, chat bad”. The point is diligence: the same technology, but with follow-ups, source control, data protection and traceability — the small things that, in law, are anything but small. How we frame this is on Why not just ChatGPT? and the AI transparency page.
The fastest way to see it is a direct comparison: ask your next question once in the Guided mode of Legal Questions — and once in a bare chat.
