Insights
Research & Analysis
Original research on litigation intelligence, behavioral game theory, and AI-powered evidence management.
New here? Start with Sentient Analysis for the thesis, then Red Team Your Case for the scoring framework, then Adversarial Consensus for how we validate.
Market Research
The Litigation Dead Zone
86% of civil trial winners receive awards under $250,000. 5.6 to 9 million cases per year fall in a range that institutional funders refuse to touch. AI changes the math.
Thesis
Complexity as Signal: Why AI-Augmented Litigation Works Differently
In a classically staffed case, complexity is a weakness. In an AI-augmented case, complexity is the primary source of value. The connections between cases are the product.
Safety Engineering
Safety Shield: Physical Safety as a Constraint Layer
Safety should be a constraint, not a checklist. S-ARPN scoring integrates physical safety into litigation intelligence as a continuous constraint on the decision optimizer.
Featured Research
From Sentiment to Sentient: A New Framework for Litigation Intelligence
Every tool reads sentiment. Acquit.ai models the sentient. How modeling opposing parties as cognitive agents with belief architectures and behavioral modifiers achieves 91% prediction accuracy.
Practitioner Guide
Red Team Your Case: The ARPN Framework
Adapted FMEA for litigation intelligence. A five-factor risk scoring system that models behavioral amplification and cascade effects, with a worked example from a $32M Texas verdict.
Methodology
The Disagreement Is the Feature: Adversarial Consensus Scoring
How the Acquit Score™ is calculated. Multi-model adversarial scoring gives you multiple perspectives, and highlights the blind spots your opponent will exploit.
Research Brief
The Real Cost of eDiscovery
RAND's $1.8M median benchmark, current vendor pricing, the access gap, and a clearly labeled modeled comparison for what compression can look like.
Research Paper
Behavioral Modifiers in Litigation Game Theory
When game theory meets human reality: a framework for predicting what litigants will actually do. Visual explainer with interactive payoff models.