How legal AI is used
Prompt Type Distribution
What lawyers ask AI to do, split across research, drafting, analysis and more.
HAQQ-measured prompt data shows legal work concentrated in two categories: Legal Research & Analysis (24.3%) and Document Drafting & Review (21.7%) together account for nearly half of all prompts, with Contract Analysis a distant third at 14.2%. Litigation Support (2.4%) and general questions (1.8%) sit at the bottom, suggesting AI adoption in legal practice is still anchored in research and drafting rather than courtroom-facing work. For legal teams evaluating AI tools, this points to research and drafting quality as the highest-leverage areas to benchmark first.
Data
| Prompt type | Share |
|---|---|
| Legal Research & Analysis | 24.3% |
| Document Drafting & Review | 21.7% |
| Contract Analysis | 14.2% |
| Regulatory Compliance | 11.8% |
| Case Strategy & Argumentation | 8.5% |
| Legal Summarization | 6.9% |
| Due Diligence | 4.8% |
| Client Communication | 3.6% |
| Litigation Support | 2.4% |
| Other / General Questions | 1.8% |
Naměřená data, předzpracovaná ze skutečného datového souboru o využití HAQQ Legal AI Index (více než 134 000 datových bodů z 30 zemí) nebo z vlastního právního benchmarku HAQQ pro více modelů hodnoceného z 50.
Související výzkum
- Top Legal Tasks by AI UsageThe legal tasks lawyers most often hand to AI, ranked by share of usage.
- Legal AI Request ClustersThe largest clusters of real legal requests sent to AI, by volume.
- Prompt Engineering ImpactHow prompting technique changes accuracy, completeness and relevance of legal answers.