# The Logic Behind Our Risk Assessment ## Why Logic Matters in Legal Risk Most compliance tools — especially those leveraging AI — are built on **propositional logic**. This means: - Binary conditions (true/false) - Logic gates (`AND`, `OR`, `NOT`) - Truth tables This structure is seductive: it's clean, computable, and makes rules feel like code. But legal risk doesn't work that way. ## From Truth Tables to Intersections Where AI compliance systems often ask: > "Is this condition true _and_ that condition true?" We ask: > "Where do systems, roles, and responsibilities **intersect**?" This is **set logic**, not propositional logic. We use `\u22C2` (the set intersection symbol) as a guiding logic. It means: - Risk emerges where domains overlap - Duties arise not from linear triggers, but from **shared membership** in legal categories - Context matters: who is involved, what systems touch the data, and what jurisdictions apply ## The Role of Thresholds Our thresholds aren't hard-coded as binary rules — they are **derived from the density of intersections**: - A threshold is reached when **enough categories intersect** - Risk is amplified not by a single truth condition, but by a **network of shared obligations** This shifts the model from: - "True AND True = High Risk" To: - "This node sits at the **convergence of high-sensitivity categories** = Elevated Risk" ## Design Philosophy - We don't use `!important` logic — we use **visible structure** - We don't flatten compliance into truth tables — we **map responsibilities as nodes** - We don’t ask “Is this compliant?” — we ask “**Where is this accountable?**” ## Why This Matters Because legal risk is not mechanical. It’s: - Relational - Contextual - Often **shared** across multiple actors This logic is not just technical — it's philosophical. And it informs: - How we define thresholds - How we build YAML properties - How we design the UX - How we help lawyers see complexity without drowning in it This is not `∧` logic. It’s `⋂` logic. And that’s the core of our toolkit