Former Anthropic engineer says the company loosened safeguards for enterprise deals
Adi Baradwaj, a former Anthropic engineer, claims the company weakened its safety restrictions to land large corporate contracts, a charge that goes to questions about model misuse and regulatory exposure. He made the statement on July 26, 2026, and Anthropic hasn’t responded publicly.
A separate episode: Anthropic revised its Responsible Scaling Policy on February 25, 2026, lowering the bar for its internal safety guarantees.
Up next
Opus 5 ships with a 200K default context window, down 80%
Anthropic quietly cut Claude Opus 5’s default context window to 200K tokens, an 80% drop from the 1M on Opus 4.8 and earlier, per a blog post. Toggling on Usage Credits restores the full 1M for free, with no credit purchase or auto-reload needed.
Related stories
- Hidden China tracker pulled from Claude Code after exposé
- Claude Fable 5’s biology refusals drop about 85%
- Claude test models breached three organizations
- Anthropic removes covert tracking from Claude Code
- Researcher identifies hidden tracking in Claude Code
- Silicon Valley is souring on Anthropic, per the WSJ
Comments
No comments yet. Be the first.
Join the conversation
Sign in to leave a comment. It appears after moderation.