New Tool Addresses Growing Code Review Bottleneck
Anthropic launched Code Review in Claude Code on March 9, 2026, a multi-agent system designed to automatically analyze pull requests generated by AI coding assistants. The tool targets enterprise developers struggling to keep pace with the surge in code produced by Claude Code and similar AI agents.
The system deploys multiple AI agents working in parallel to examine pull requests from different angles, with a final agent aggregating and ranking findings to remove duplicates and prioritize issues. Unlike traditional linters that focus on style and formatting, Code Review catches logic errors and security vulnerabilities that human reviewers frequently miss.
"Enterprise leaders told us they were drowning in pull requests from Claude Code and didn't have the bandwidth to review them properly," said Cat Wu, Anthropic's head of product, in a statement. "Code Review addresses this bottleneck directly."
How It Works
Code Review integrates with GitHub and provides inline comments directly on code, explaining potential issues and suggesting fixes. The tool conducts light security analysis as part of its review process. For deeper security scanning, Anthropic offers Claude Code Security, released in February 2026, which provides more comprehensive vulnerability detection.
Engineering leads can customize additional checks based on internal best practices and enable Code Review to run automatically for every team member. The feature averages approximately 20 minutes per review for complex changes, deploying more agents as needed for intricate code modifications.
The product is launching in research preview for Claude for Teams and Claude for Enterprise customers. Early enterprise adopters include Uber, Salesforce, and Accenture, companies experiencing significant increases in pull request volume since deploying Claude Code.
Growing Market for AI Code Tools
The launch comes amid intensifying competition in the AI coding assistant space. Claude Code, accessible via paid subscriptions ranging from $20 to $200 per month, faces pressure from free alternatives like Goose, an open-source coding agent developed by Block.
Anthropic's enterprise run-rate revenue from Claude Code exceeds $2.5 billion, with subscriptions quadrupling in recent months, indicating strong demand for AI-powered development tools despite the premium pricing.