Anthropic is dropping its signature safety pledge amid a heated AI race
Business Insider
February 25, 2026
AI-Generated Deep Dive Summary
Anthropic, a leading AI startup known for its commitment to safety, has announced a significant shift in its foundational approach. The company is no longer adhering to its previous pledge to pause or delay the deployment of new AI models when advancements outpace their safety measures. This decision comes amid heightened competition and a lack of government regulation in the rapidly evolving AI landscape. Anthropic CEO Dario Amodei has emphasized the need for adaptability, stating that unilateral commitments to pause model development no longer make sense if competitors are advancing quickly.
The company is introducing a new framework called its Responsible Scaling Policy, which draws inspiration from the US government's biosafety level (BSL) standards. While this policy still includes a commitment to delay the release of highly capable AI models in certain circumstances, it provides more flexibility and aligns with the current state of the AI race. Anthropic’s chief science officer, Jared Kaplan, explained that the old policy was not in line with the rapid pace of innovation, as competitors are "blazing ahead." The company also cited an anti-regulatory political climate as a key factor in its decision to revise its approach.
Amodei has long championed AI safety, even quoting Spider-Man’s mantra, "With great power comes great responsibility," to highlight the risks associated with advanced AI. However, he acknowledged that some safety measures could no longer be contained by a single company. Anthropic’s decision also reflects the challenges it faces in balancing innovation with public trust, especially as its flagship chatbot, Claude, continues to disrupt financial markets and raise concerns about the future of software.
The company’s move away from its original safety pledge underscores the pressures faced by AI startups navigating an unregulated environment. While Anthropic has pushed for state-level AI regulations, federal efforts have been slow to progress. Amodei remains committed to advocating for effective government engagement on AI safety but acknowledges it as a long-term project. The revised policy, described as a "living document," aims to adapt to the evolving risks and capabilities of AI while maintaining a focus on responsible development.
This shift in Anthropic’s strategy matters to business readers interested in understanding how companies are balancing innovation with caution in the AI race. By revising its safety policies, Anthropic is signaling a pragmatic approach to staying competitive while addressing the limitations of unilateral commitments. The company’s decision highlights the broader challenges
Verticals
businessfinance
Originally published on Business Insider on 2/25/2026