Trust, risk & security
Client Trust & Satisfaction
Trust, satisfaction and reuse rates for AI across legal work products.
As a directional estimate, trust in AI output tracks how structured the task is rather than how "legal" it is. Compliance Checklists (80 trust, 90% would-reuse) and Document Review Flags (75 trust, 88% would-reuse) score highest, while Client Communications sits lowest at 45 trust and just 48% would-reuse. The gap suggests lawyers are comfortable delegating rule-based, checklist-style review to AI but remain wary of letting it touch judgment-heavy or client-facing work, where a wrong call carries relationship risk, not just an error to catch.
Los datos
| Category | Trust | Satisfaction | Would reuse |
|---|---|---|---|
| Legal Research Summaries | 72% | 78% | 85% |
| Contract First Drafts | 58% | 65% | 72% |
| Compliance Checklists | 80% | 82% | 90% |
| Case Law Analysis | 64% | 70% | 76% |
| Document Review Flags | 75% | 80% | 88% |
| Client Communications | 45% | 52% | 48% |
| Regulatory Filings | 50% | 58% | 62% |
| Due Diligence Reports | 55% | 62% | 68% |
Estimación ilustrativa - una cifra orientativa para la configuración de escenarios, no una referencia medida. No los interprete como resultados medidos por proveedor.
Investigación relacionada
- Risk & Hallucination RatesEstimated error rates across legal tasks, highest for citation and jurisdiction work.
- Tasks That Still Need a HumanThe share of each legal task that professionals say still requires human judgment.
- Data Security & CompliancePublicly documented security certifications across major AI providers.