Round seven of the markup, and nobody can remember who moved that clause

Anyone who has run a negotiation past round four knows the feeling.

Three firms are marking up the same agreement. Two of them are accepting changes as they go. Somebody sends a clean copy that is not actually clean. Someone else starts working from an older version. A client-side lawyer makes three changes in a document that comes back looking almost identical to the one before it.

And then, on Thursday afternoon, the partner asks a question that should be simple: When did the indemnity cap move from twelve months to eighteen, who moved it, and did we ever agree to it?

The answer is usually buried somewhere in a folder containing eleven versions of the same agreement. Finding it means opening documents, comparing versions, following tracked changes and remembering which version came before which. It is not intellectually difficult work. It is reconstruction work.

And reconstruction is becoming a bigger problem. The final agreement tells you where the negotiation ended. It does not always tell you how it got there.

For a long time, the practical solution was simply to keep the versions. If someone needed to know what happened, they could go back through them. But that approach assumes that a human being has the time, context and patience to reconstruct a negotiation from a pile of documents.

AI changes that possibility.

Instead of treating a redline as a visual document that needs to be read from top to bottom, it can be treated as a record of decisions. Changes can be separated from formatting. Substantive movements can be distinguished from cross-reference renumbering. Changes can be grouped by author and traced to the exact language that was altered.

That is the direction we have been exploring at Jurisphere with Redline Analysis.

The interesting part is not that it can tell you that a clause changed. A lawyer can see that. The interesting part is what happens when hundreds of individual changes become a structured history of the negotiation. You can ask what changed, who changed it, when it changed and which changes were actually material.

That turns the redline from a document you inspect into a record you can interrogate.

Legal work has always produced a trail of decisions. We have simply been poor at making that trail usable. The information exists across drafts, emails, comments and markups, but reconstructing it has traditionally been too expensive in lawyer time to do routinely.

Two years after a transaction closes, nobody wants to discover that the answer to an important question is sitting somewhere in “SPA_v7_FINAL_clean_revised2.docx”.