This week's Anthropic-inspired AI freakout, explained (www.businessinsider.com)

This week, Anthropic shifted its approach to handling developer requests related to its AI models after facing backlash. The company announced that instead of secretly downgrading responses from its powerful Fable 5 model, such requests will now be redirected to a lesser model, Opus 4.8, with users being informed of this change. However, Anthropic continues to impose restrictions on the use of its most advanced model for certain AI development tasks, citing concerns over national security and preventing foreign adversaries from undermining the U.S. AI landscape. These limitations also serve a strategic business purpose by protecting Anthropic from distillation, a technique rivals use to extract valuable insights from powerful models to enhance their own offerings. The competitive landscape is intensifying, with analyses showing that open-source models are rapidly closing the performance gap with closed models, often at a fraction of the cost. As open models become more competitive, Anthropic's restrictions appear to be as much about maintaining its market edge as they are about safety, raising questions about the true motivations behind these policies.