Latent Space AI Anthropic Launches "Code Review" to Fix AI Code Security Issues
Mar 9, 2026
They discuss a new AI tool that scans generated pull requests for logical bugs and security holes. The conversation covers how rising AI-assisted coding is creating review bottlenecks for teams. Listeners hear about the tool's multi-agent analysis, severity labels, and where deeper security audits and pricing come into play.
AI Snips
Chapters
Transcript
Episode notes
AI Code Is Creating A Review Bottleneck
- AI is producing a massive volume of code and that creates a new bottleneck in reviewing pull requests.
- Jaeden Schafer cites estimates that 70–90% of some companies' code is AI-generated, making manual review impractical.
OpenClaw Founder Got Flooded With Pull Requests
- Jaeden recounts the OpenClaw viral episode where the solo maintainer was overwhelmed by incoming PRs.
- The founder said he was bogged down trying to review community submissions after the project went viral and was later acquired.
Open Source Hesitates While Enterprises Embrace AI Code
- Some open-source communities ban AI-generated code because the flood of PRs makes reliable review hard.
- Jaeden notes enterprises like Uber and Salesforce are already heavy Cloud Code users, driving demand for automated review.