Anthropic's Fully Automated Code & Security Review Raises Alignment Concerns
ronbodkin · x · 2026-07-21
AI engineer @ronbodkin raised concerns over Anthropic's internal "Stage 3" development workflow. The process completely removes human review in favor of fully automated guardrails, which is disturbing given that control and alignment remain unsolved problems.
Anthropic's current automated stack includes:
- Automatic code and security review
- Agent sandboxing
- Using CLAUDE.md and Skills to encode standards
- Tuning Auto mode classifiers based on team usage
- Managing token consumption via model selection, advisors, LSPs, and breaking down CLAUDE.md into lazy Skills
More from coding & agent
- Reticle builds a verification layer that lets AI coding agents check their own generated code — Scobleizer · 2026-10-27
- Trust your agent, but read its logs: Claude managed agents tips — brada · 2026-10-27
- Dev training a guard model to block prompt injection, tool and memory poisoning — gnukeith · 2026-10-27
- Dev trains a ~4GB-VRAM sentinel model to guard against prompt injection and poisoned memory — gnukeith · 2026-10-27
- Superset rebrands as Paradev, raises $11.5M seed led by USV — massimosgrelli · 2026-10-27
- Ex-Big Tech managers may thrive in the AI era: manage agents like you managed teams — lxfater · 2026-10-27