The short answer: most legal AI works at the level of a single document. The real leverage is at the level of the matter, the whole file: the emails, the documents, the history and the live law. When the AI holds the matter, every task starts with the full picture instead of a blank page.
Document level AI, and its ceiling
A lot of legal AI is very good at one document. Tools like Spellbook draft and review a contract inside Word with real skill. Platforms like Legora can read many documents at once and return the answers in a structured grid. Both are genuinely useful. But a single contract, or even a grid of contracts, is not the same as the state of the matter. As Spellbook’s own scope makes clear, a contract tool understands the contract, not the matter behind it.
What the matter actually contains
A matter is more than its documents. It is the client’s instructions, the chain of emails, the deadlines, the prior drafts, the decisions already made, and the live law that applies. A lawyer carries that context in their head. Most AI does not, which is why it has to be re-briefed for every task.
Why matter context changes every task
When the AI holds the matter, the work compounds:
- Research starts from the facts of this matter, not a blank search box.
- Drafting begins with everything the file already knows, in your firm’s voice, rather than a generic template.
- Review is measured against what this deal is meant to achieve, not just a checklist.
- The reply to the client is written knowing everything that happened, because the system was there for all of it.
Qanooni is built around the matter for exactly this reason. It connects your documents, your correspondence and the live law into one place, so the AI works the way a lawyer does: with the whole file in view, not one page at a time.