Responsible Scaling Policy
A responsible scaling policy is a developer's public commitment to tie the scaling up of its AI models to specific safety and security measures triggered when evaluations show dangerous capability levels.
What is Responsible Scaling Policy?
Anthropic introduced its RSP in 2023; OpenAI's Preparedness Framework and Google DeepMind's Frontier Safety Framework are similar.
How it works
Capability thresholds are defined in advance; models are evaluated and stronger safeguards apply before training or deploying beyond each threshold.
← All Artificial Intelligence & Superintelligence definitions · ← Back to the A–Z definitions