Anthropic's Core Safety Pledge Now Optional, Company Says 'Competitive Pressure Is The New Reality'
SAN FRANCISCO — Anthropic, the Silicon Valley darling that once promised to halt model development if safety procedures outpaced innovation, has officially dropped its flagship safety commitment. The company now describes this decision as “strategic optimization” while simultaneously running full-scale ad campaigns to remind you that their chatbots have a “higher duty of care.”
The Responsible Scaling Policy Version 3.0 was announced Tuesday, effectively removing the company’s previous guarantee that it would pause model releases if safety protocols couldn’t keep up with deployment timelines. “We’ve learned that safety-first isn’t a business model, it’s a competitive disadvantage,” said Daniel Rockmore, Anthropic’s VP of Strategy, who later was photographed wearing a $42 safety-conscious t-shirt that reads “I Pause For Safety” while standing next to a banner announcing the policy change.