How Tech Companies Can Make Recent AI Commitments Count

David Morar, PhD is Senior Policy Analyst and Prem Trivedi is Policy Director of New America’s Open Technology Institute. U.S. President Joe Biden and Vice President Kamala Harris meet with leaders of Amazon, Anthropic, Google, Inflection, Meta, Microsoft, and OpenAI at the White House, July 21, 2023. Source Last week, U.S. President Joe Biden announced alongside CEOs of leading AI companies that the White House had secured “voluntary commitments” from seven companies to build safe and responsible AI systems. This announcement is proof of the Biden administration’s continued focus on managing both the promise and the perils of artificial intelligence. And participating companies deserve some credit for making reasonably detailed statements of intent. The press release and accompanying statement of commitments emphasize that the document is “an important first step” toward the development of laws and rules.  That acknowledgment is important, but it is equally important that such a first step produces meaningful action on its own. On that front, much work remains. Building an effective voluntary commitments process requires incorporating inputs from civil society into the lifecycle of companies’ internal governance process. It also requires identifying specific timelines for the deliverables, along with establishing auditing and monitoring mechanisms. Amazon, Anthropic, Google, Inflection, Meta, Microsoft and OpenAI agreed to eight commitments across three categories of principles — safety, security and trust. The commitments include internal and external security testing, information sharing on AI risks, investments in cybersecurity safeguards around model weights, third-party discovery and reporting of vulnerabilities, robust technical mechanisms…How Tech Companies Can Make Recent AI Commitments Count