A court-accepted eDiscovery methodology using machine learning to rank documents by relevance, reducing manual review volume; also called CAL or CAR.
Last reviewed: 2026/05/19
A court-accepted eDiscovery methodology using machine learning to rank documents by relevance, reducing manual review volume; also called CAL or CAR. Define technology-assisted review, then connect e-discovery readers to workflow pages, tools, and comparisons.
An iterative ML approach in eDiscovery where the model continuously updates relevance predictions as reviewers code documents, prioritizing the most uncertain documents for review.
SecurityUsing AI to identify, notify custodians, and track preservation obligations when litigation or investigation triggers a duty to preserve electronically stored information.
CapabilityA TAR technique where the system learns from attorney-coded seed documents to predict relevance across the full document set; court acceptance depends on validation methodology.
Move from this definition to role-based legal AI shortlists and the selection criteria that matter for each type of legal team.
Am Law 200 and global firm workflows: accuracy at scale, security compliance, and matter-level auditability.
Last reviewed: 2026/05/19. Definitions are written by the LawyerAI Editorial team. Commercial relationships are disclosed and do not determine editorial scores or conclusions. See our Sponsorship & Affiliate Disclosure.
Technology-assisted review (TAR) is a court-accepted eDiscovery methodology that uses machine learning algorithms to prioritize, rank, or classify documents by predicted relevance, substantially reducing the volume of documents requiring manual attorney review. Also referred to as computer-assisted review (CAR) or continuous active learning (CAL) when using iterative training methods, TAR learns from attorney coding decisions and applies those decisions to the broader document set. Courts have accepted TAR as an appropriate review methodology when parties agree on protocols and validate outcomes through recall and precision sampling.
Document review is the largest cost component of civil litigation, consuming 70-80% of eDiscovery budgets in large cases. A large commercial dispute may involve millions of documents; manual review of every document is economically prohibitive. TAR addresses this by focusing attorney review time on the documents most likely to be relevant, substantially reducing review volume without proportionally reducing recall.
The court acceptance history of TAR is well-established. The Da Silva Moore, Rio Tinto, and Progressive Casualty cases established federal court acceptance of TAR methodology in the early 2010s. Courts generally accept TAR when parties negotiate protocols in advance, document the training process, and validate outcomes through agreed sampling.
Lawyers using TAR must understand the methodology sufficiently to defend it. Opposing counsel may challenge TAR protocols — questioning seed set selection, training document quality, or recall validation sampling. The supervising attorney must be able to explain and defend the TAR process, not merely delegate it entirely to a vendor.
Relativity is the dominant eDiscovery platform for large-scale TAR, with its Active Learning module implementing continuous active learning that prioritizes documents for review as reviewers code, continuously updating predictions. DISCO integrates TAR within its cloud-native eDiscovery platform, with visualization tools that show review progress and estimated remaining relevant document populations.
Logikcull offers TAR functionality with a more accessible interface designed for smaller matters and less experienced users, reducing the technical barrier to TAR adoption in mid-market litigation.