AI Breakdown Anthropic Launches "Code Review" to Fix AI Code Security Issues
Mar 9, 2026
A deep look at a new AI code review tool that checks AI-generated pull requests for bugs and security risks. Discussion of how automated reviews integrate with GitHub and leave human-like comments. Examination of multi-agent analysis, severity labeling, and the limits of the tool's security checks. Notes on customization, pricing, and potential effects on engineering workflows.
AI Snips
Chapters
Transcript
Episode notes
Actively Scan AI Generated Code For Hidden Risks
- Don't assume AI-generated code is safe; actively scan for hidden bugs and security risks.
- Jaeden warns AI tools produce code developers may not fully understand, creating hidden vulnerabilities that need review.
AI Generated Code Is Dominating Engineering Workflows
- AI is rapidly generating a huge share of code, creating volume that outpaces traditional human review.
- Jaeden Schafer cites estimates where 70–90% of some companies' code is AI-generated, creating a new review bottleneck.
Enable Automated PR Reviews For Every Pull Request
- Use automated review tools that run on every pull request to catch issues before merging.
- Jaeden highlights Anthropic's Code Review in Cloud Code which integrates with GitHub and leaves comments directly on PRs.