Anthropic is pushing Congress to mandate independent safety tests for powerful AI models

Quartz Quartz

The company wants independent evaluations required for models trained above a compute threshold, with government authority to block dangerous deployments

Read full article at Quartz →