🛡 Anthropic Scraps Policy of Secretly Throttling Claude Models
Anthropic has withdrawn the controversial practice of using Claude Fable 5, which involved covertly reducing performance when users attempted to engage in advanced AI model development. Instead of "invisible sabotage," the developers have promised to make protective mechanisms transparent: users will now receive notifications when queries are blocked or redirected to less powerful models.
🌍 This decision highlights the growing tension between proprietary model developers and the research community.
👤 This means that using Claude for deep technical research will become more predictable.