Responsible Use
Many attorneys and self-represented litigants use GenAI tools to help with research and drafting, as well as organizing case materials and presenting evidence. Some judges use GenAI to facilitate their work. This section explores four judicial use paradigms: legal research; drafting and editing; summary; and other (mostly non-judicial) tasks. Underlying each of these paradigms is the core obligation to ensure that independent judgment, reliability, and confidentiality are not compromised.
The Federal Judicial Center is neither endorsing nor discouraging the use of GenAI in chambers. The material in this section is intended to provide judges with information to consider as they independently assess whether to integrate GenAI into their workflow.
-
Legal Research
GenAI can support – but not replace – traditional legal research. Legal AI systems tend to be more reliable than general purpose models, but they are not without risk of hallucinations and other errors. Human verification is always necessary.
-
Drafting and Editing
GenAI can facilitate a broad range of drafting and editing tasks. As with other AI-use paradigms, there are strategies and protocols for using GenAI drafting assistance responsibly, including selecting lower risk drafting tasks and not outsourcing analysis and judgment to AI.
-
Summarization
GenAI can assist with summarizing and organizing documents. However, these tasks may involve analysis or choices by the AI tool and may produce incomplete or inaccurate documents. Judges should assess which tasks are appropriate for GenAI assistance and review output.
-
Other Tasks
Judges may use GenAI to support a variety of administrative and other tasks, for example, organizing email, scheduling, training chambers staff, and preparing materials for speaking engagements. Even when used for non-judicial work, GenAI output should be verified.