Anthropic, a company initially dedicated to AI safety, has revised its Responsible Scaling Policy (RSP) amid heightened competition and the absence of government regulations. Founded by former OpenAI employees, Anthropic is now altering its safety measures due to the evolving landscape of AI development.
The company’s updated policy eliminates the commitment to pause or delay the scaling of new AI models if their advancement surpasses safety controls. This change reflects a strategic shift towards prioritizing AI competitiveness and economic growth over stringent safety protocols.
Anthropic’s Chief Science Officer, Jared Kaplan, acknowledged that the previous safety policy failed to keep pace with the rapid advancements in AI technology. Kaplan emphasized the need to continue training AI models despite the potential risks, citing the competitive pressure from other industry players.
This decision underscores the evolving dynamics within the AI sector, where companies like Anthropic are compelled to recalibrate their safety strategies to remain competitive in an environment lacking regulatory oversight.
Source: mint – technology