Pacing would let interpretability research catch up to model capability

Artificial intelligence research, products, and impact across society.

The field of preventing harm from advanced AI systems, spanning alignment, evaluation, and decisions to slow or withhold models whose capabilities outpace their safeguards.

The September 2026 debate sparked by Anthropic CEO Dario Amodei's essay "We Must Pace the Frontier", which argues that AI development should be deliberately slowed so that safety, alignment and interpretability can keep pace, and sets out a three-step plan spanning company commitments, coordination among democracies, and US-China agreements.

The scientific study and development of artificial intelligence systems, algorithms, and methodologies.

Methods for understanding and explaining how models make decisions, including predictions and internal representations. Supports transparency and debugging.