← All stories
● Covered by 1 source · 1 reportMedium impact1 neutral

Uber's Legal AI Agent Reduces Contract Review Time by Over 20%

🔄 Updated 5d ago
New to BrevFeed? We gather this story from every outlet covering it into one summary — ranked by real-world impact, not just the latest headline — so you never miss what matters. What is BrevFeed? →

Key points

  • Uber's AI agent cut contract review time by over 20%.
  • The agent achieves 91% accuracy in AI-generated decisions.
  • It uses RAG, historical feedback, and rules for recommendations.
  • Lawyers retain final approval over AI suggestions.

AI Streamlines Contract Review at Uber

Uber's Legal Redlining Agent, an AI tool, has reduced the average time spent on contract review by more than 20%. The system operates with a 91% accuracy rate in its AI-generated decisions. This agent is integrated into Microsoft Word, where it analyzes proposed contract changes from external parties.

Functionality and Implementation

The agent assesses legal risks and recommends policy-aligned responses. It utilizes retrieval-augmented generation (RAG), historical lawyer feedback, and deterministic rules to formulate its suggestions. Lawyers are responsible for reviewing and approving the agent's recommendations, ensuring human oversight in the process.

The tool identifies edits, interprets their intent, compares changes against legal policies, and suggests whether to accept, reject, or modify clauses. It also generates comments to explain recommendations and flags changes requiring closer scrutiny. Uber began developing the first generation in early 2024 and piloted it with its Legal team in 2025.

Evolution of the AI System

Initially, a RAG implementation retrieved relevant sections from legal playbooks. However, this approach faced challenges with semantic retrieval accuracy, negotiation tone, and handling scenarios not covered by existing examples. Uber addressed these issues by adding a feedback loop that captures lawyers' decisions, reasoning, and final edited text.

This feedback loop allows the agent to retrieve examples of how similar clauses were handled previously, rather than just whether a recommendation was accepted or rejected. At runtime, similarity search and metadata filtering narrow historical interactions to approximately 20 candidates, which are then evaluated for relevance by a second language model. An exponential decay weighting, with a 365-day half-life, prioritizes recent decisions while retaining older examples. Domain experts also control language and negotiation style.

Technical Architecture

The system's architecture combines a React and TypeScript Word add-in with a Python backend orchestrated through LangGraph. Parallel workflow nodes manage intent detection, risk assessment, and policy lookup. OpenSearch stores vectorized legal rules and historical feedback, while Redis maintains session state for follow-up questions.

✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →

The daily brief

One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.

One email a day. Unsubscribe in one click, any time.

Today's brief

Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.

~4 min · 3 stories · Nov 21

▶ Play today's brief Listen on Spotify

New every morning, and the back catalogue is archived by date.

Reporting from

Uber developed and deployed a Legal Redlining Agent, an AI tool integrated into Microsoft Word, which reduced average contract review time by over 20% with 91% accuracy. The agent analyzes proposed contract changes, assesses risks, and recommends responses using RAG, historical lawyer feedback, and deterministic rules. This system allows lawyers to maintain oversight while automating parts of the contract review process.