Ask most litigators where they lose time, and they'll tell you: research, drafting, document review. These are the obvious answers, and they're not wrong. They're also the answers that the current generation of legal AI has been built to address.
But there's a more fundamental place where litigation work breaks down — one that doesn't show up on any timesheet and doesn't get solved by faster research or better drafts. It's the work of holding the case together.
The Invisible Tax on Every Matter
Before a litigator can do any substantive work on a case, he or she has to understand the case. That means knowing the key facts, the contested issues, the prior rulings, the operative theory, and how the adversary is likely to respond to any given move. On a matter that has been running for six months, with a team of eight and a document set in the hundreds of thousands, getting to that understanding takes time, and then has to be repeated every time a team member changes, every time strategy shifts, and every time something significant happens in the case.
This is the invisible tax that litigation teams pay on every matter. It doesn't appear as a discrete line item. It appears as hours spent in status meetings, as duplicated research, as junior work product that has to be redone because the associate didn't have the full picture. Or a partner who has to reconstruct the case theory before she can review a brief.
It compounds. The longer a matter runs and the more complex it becomes, the higher this tax gets. And no existing tool—not eDiscovery software, not legal research platforms, not general-purpose AI assistants—has been built to address it.
Where the Breakdowns Actually Happen
There are specific moments in a litigated matter where the absence of a continuous case understanding is most costly. A few of the most common:
Team transitions. When an associate leaves a matter or a new one joins, the incoming person starts from zero. In theory, a handoff memo bridges the gap. In practice, it captures a fraction of what someone needs to know, and it's already out of date by the time it's written. The new associate spends days or weeks reconstructing context that existed somewhere in the team's collective memory — but not in any system.
Depositions. Deposition preparation requires a complete command of everything the witness has said, everything the documents show, and every inconsistency worth surfacing. In practice, this preparation is exhausting and imperfect. During the deposition itself, there's no live system checking testimony against the full record. If a witness says something that contradicts a document from 14 months ago, catching that contradiction depends on who's in the room and how much they remember.
Motion practice. A brief argues a theory. The theory should be grounded in the current state of the record and the current strategic posture of the case. In practice, the connection between the case's living strategy and the document a junior associate is drafting is often tenuous. The associate knows the assignment; she may not know how it fits into everything else.
Strategy shifts. Cases evolve. Facts that seemed dispositive turn out not to be. Rulings create new constraints. Witness testimony changes the picture. When strategy shifts, the entire team's work product has to be reconsidered in light of the new direction. This happens informally, imperfectly, and often incompletely — because there's no system that represents the case as it currently stands.
Why Existing Tools Don't Close These Gaps
The legal AI tools available today are powerful at the task level. They surface case law, draft motions, review documents at scale. Point solutions go narrower, targeting specific workflows like deposition prep or document review. But narrower isn't deeper.
None of them hold the case. They don't know what happened in last week's deposition, that the theory shifted after the last ruling, or that a document produced two months ago changes how everything else reads. They answer a question and start fresh next time.
A Different Architecture for a Different Problem
Closing the gaps described above requires a different starting point. Not a task-level AI that answers questions, but a case-level system that maintains a continuous understanding of the matter as it develops.
This means structuring and preserving everything that defines the case: documents, rulings, depositions, strategy, evolving theories, and prior work product. It means building an intelligence layer that compounds over time, so that what happened in month three is still available and connected to what's happening in month eighteen. It means operating at the matter level, not the query level.
When the system understands the case, the visible efficiencies follow. Drafts are grounded in the current record. Team transitions don't reset the clock. Deposition prep draws on the full picture. And the invisible tax — the hours spent reconstructing context that shouldn't need to be reconstructed — starts to come down.
That's the breakdown worth solving. It's not the most visible one in litigation. But it's the one that everything else runs through.
Advocacy is an AI-native litigation technology company building the first context infrastructure for legal teams. Request an introduction and case evaluation here.