Relativity aiR in brief
Relativity aiR is Relativity's generative AI suite for legal data intelligence inside RelativityOne. It targets litigation, investigations, document review, privilege work, and case preparation. Relativity announced limited availability for aiR for Review at Legalweek 2024, followed by general availability for Review in September 2024 and Privilege in November 2024. The company reported more than 100 million aiR document decisions during the 20 months ending in October 2025. Relativity also reported $160 million in research and development investment for 2025, a company wide figure rather than an aiR budget. The suite is designed for legal teams that need review controls, auditability, and AI assistance within an existing case data system. Its positioning therefore emphasizes controlled deployment within established litigation workflows rather than replacement of attorneys or outside counsel.
Relativity aiR begins with extracted text and images in a RelativityOne workspace. In aiR for Review, a user writes natural language criteria, tests prompts on a smaller subset, and analyzes a collection. Results identify significant documents, explain their relevance, and cite supporting passages. Privilege analyzes documents and relevant personas, predicts privilege, and drafts privilege log descriptions. Case Strategy supports fact extraction, timelines, summaries, and work product. Assist provides conversational search over legal data. Outputs include rationales and considerations, while Review Center can calculate precision, recall, and elusion metrics for quality control. These controls make the system more suitable for defensible review than a general purpose chatbot operating outside the matter workspace. They also give review leaders a structured basis for sampling, escalation, and documenting decisions before production.
The main limitation is that aiR depends on RelativityOne, which makes it less suitable for teams choosing a standalone review system. Inference may occur outside the instance storage region when Azure model availability requires it, although Relativity says customer data is not used to train third party models. Everlaw offers a separate cloud discovery platform with comparable AI review workflows, while DISCO focuses on cloud discovery and case preparation without requiring RelativityOne. Clearbrief focuses on cited drafting and discovery-related writing, and Briefpoint supports automated legal drafting as an add-on product. Westlaw Precision is aimed at legal research rather than matter data. Choose aiR when review, privilege, validation, and case intelligence should remain in one RelativityOne workspace. Confirm regional processing, contract scope, and migration costs before standardizing on it.