The Biden administration on Friday announced it had received voluntary commitments to institute safeguards around the technology from seven firms, including Amazon, Meta and Microsoft as well as startups Anthropic and Inflection. (The Logic)

Talking point: The firms agreed to let “independent experts” security-check their AI systems before launch, a step down from the licensing regime that OpenAI CEO Sam Altman, the sector’s loudest voice on regulation, has proposed. The companies have also committed to publicly say what their platforms can, can’t and shouldn’t do, as well as to develop technology to address climate change, health care and other challenges. The Biden administration’s opt-in, eight-point plan comes as U.S. policymakers and industry try to shape AI regulation and legislation. It’s an interim step to another interim step: the U.S., EU and Canada are jointly drafting a voluntary code for AI companies to bridge the gap to new laws governing the technology.