For our upcoming AI+Access to Justice Summit and our AI for Legal Help class, our team has made a new design workbook to guide people through scoping a new AI pilot.
We encourage others to use and explore this AI Design Workbook to help think through:
- Use Cases and Workflows
- Specific Legal Tasks that AI could do (or should not do)
- User Personas, and how they might need or worry about AI — or how they might be affected by it
- Data plans for training AI and for deploying it
- Risks, laws, ethics brainstorming about what could go wrong or what regulators might require, and mitigation/prevention plans to proactively deal with these concerns
- Quality and Efficiency Benchmarks to aim for with a new intervention (and how to compare the tech with the human service)
- Support needed to go into the next phases, of tech prototyping and pilot deployment
Responsible AI development should be going through these 3 careful stages — design and policy research, tech prototyping and benchmark evaluation, and piloting in a controlled, careful way. We hope this workbook can be useful to groups who want to get started on this journey!