OpenAI Supports U.S. House Bill Mandating External AI Safety Evaluations

OpenAI has supported a bipartisan U.S. House bill mandating external safety evaluations for frontier artificial intelligence developers, according to reports. Chris Lehane, OpenAI's Global Business Lead, reportedly endorsed the key provisions of the Frontier (FRONTIER) Act and discussed operational details with lawmakers regarding independent verification mechanisms.

The legislation, introduced on July 23, requires major frontier AI companies to establish safety frameworks, undergo independent evaluations, and report significant safety incidents. The bill's sponsors emphasized ongoing safety validation through independent assessment bodies as a core component of the proposed framework.