AI models can be dangerous before public deployment - METR
metr.org · 1,316 words · saved by 3 readers
Why pre-deployment testing is not an adequate framework for AI risk management
Many frontier AI safety policies from scaling labs (e.g. OpenAI’s Preparedness Framework, Google DeepMind’s Frontier Safety Framework, etc.), as well as past work by third party evaluators including UK AISI, Apollo Research, and METR, focus on pre-deployment testing – ensuring that the AI model is safe and that the lab has sufficient security before the lab deploys the model to the public. Such pre-deployment safety evaluations are standard for a wide variety of products across many industries, where the primary risk of the product is to the consumer (see, for example, the crash testing…
saved by
related reading
- Common Elements of Frontier AI Safety Policiesmetr.org
- Using Dangerous AI, But Safely?youtube.com
- The case for ensuring that powerful AIs are controlled — LessWronglesswrong.com
- A Safe Path to Open Weights - Thinking Machines Labthinkingmachines.ai
- AI safety is not a model propertysubstack.com
- The case for ensuring that powerful AIs are controlledblog.redwoodresearch.org
- Racing through a minefield: the AI deployment problemcold-takes.com
- AI safety is not a model propertynormaltech.ai
- Self-exfiltration is a key dangerous capabilityaligned.substack.com
- AI Safety Seems Hard to Measurecold-takes.com
- Self-Fulfilling Misalignment Data Might Be Poisoning Our AI Modelsturntrout.com
- AI safety is not a model propertyaisnakeoil.com