What happened
Anthropic, an AI safety and research company, has announced new commitments regarding the deprecation and preservation of its AI models. The company is formalizing policies to manage how older models are retired and how certain models may be preserved for research or safety purposes.
Why it matters
- Clear deprecation policies help users and developers plan for transitions when older AI models are retired.
- Preserving select models supports ongoing research, safety evaluations, and historical comparisons in AI development.
- These commitments aim to balance innovation with the need for stability and accountability in AI systems.
Key facts
- 1. Anthropic provides at least 60 daysâ notice before retiring a model, allowing users time to migrate to newer versions.
- 2. Users can audit their API usage to identify reliance on deprecated models via the Anthropic console.
- 3. Model retirement decisions are driven by the need to allocate capacity for new model releases and to address safety considerations.
