Security teams hire red teams to break their systems before attackers do. Lawyers have always done something similar, rehearsing the other side's best argument before it is made in front of a judge. What is new is software that treats that rehearsal as a built-in feature.
ilovelawyer, developed by FORHU, is an AI case workspace for lawyers, and four of its views are designed for exactly this kind of stress test. This article looks at how they work and what they suggest about useful AI in high-stakes work.
The idea: produce questions, not answers
One notable design choice is what these views refuse to do. None delivers a conclusion. Each hands the lawyer something to examine, argue with and refine. In a field where an overconfident output can cause real harm, a tool that generates challenges to the user's position is a different kind of assistant from one that generates answers.
Four views, four kinds of pressure
Red Team. It surfaces the argument the other side is likely to make and ranks likely attacks by impact. The ranking tells a lawyer where a position is most exposed, so preparation time goes to the weakest points first.
Contradictions. It flags inconsistencies across documents and testimony. The flags are for review, so the checking stays with the person.
Case Reconstruction. It writes the story of the matter in three registers: General, For the Court, and From the Other Side. Alongside it sits a plain list of its blind spots. The list marks where human effort should go next.
Theories. It keeps competing readings of the case side by side. Alternatives stay visible while the work continues, which guards against settling early on one story.