Ankündigung des Herstellerserreichbar
Anthropic, Responsible Scaling Policy: Version 3.0
Anthropics Ankündigung der dritten Fassung vom 24.02.2026, mit einer Bilanz der eigenen Theorie nach zweieinhalb Jahren. Was gewirkt hat: ASL-3 im Mai 2025 aktiviert, Open
geprüft 10.11.2026
Worauf sich diese Seite beruft, wörtlich, abgerufen am 20.09.2026:
We found pre-set capability levels to be far more ambiguous than we anticipated: in some cases, model capabilities have clearly approached the RSP thresholds, but we have had substantial uncertainty about whether they have definitively passed those thresholds.
bestätigt 10.11.2026The science of model evaluation isn’t well-developed enough to provide dispositive answers.
bestätigt 10.11.2026We activated ASL-3 safeguards for relevant models in May 2025 and have been working to improve them ever since.
bestätigt 10.11.2026within a few months of announcing our RSP, both Open
bestätigt 10.11.2026AI and Google Deep Mind adopted broadly similar frameworks. We’ve seen governments around the world (for example in California with SB 53, in New York with the RAISE Act, and with the EU AI Act’s Codes of Practice) start to require frontier AI developers to create and publish frameworks for assessing and managing catastrophic risks
bestätigt 10.11.2026