Microsoft Calls for Creation of New US Agency to Regulate AI

Microsoft has urged the US federal government to establish a dedicated agency for regulating artificial intelligence (AI), as reported by Bloomberg. During a speech in Washington, DC, attended by congressional members and non-governmental organizations, Microsoft’s Vice Chair and President, Brad Smith, emphasized that "the rule of law and a commitment to democracy have always kept technology in check" and should extend this principle to AI.

As part of Microsoft’s regulatory framework, Smith proposed the implementation of redundant AI circuit breakers—fail-safes designed to allow for the rapid shutdown of algorithms if necessary. He also recommended that President Biden issue an executive order requiring federal agencies using AI tools to follow the National Institute of Standards and Technology's (NIST) risk management framework. Smith affirmed Microsoft's commitment to these guidelines and indicated that the company would publish an annual AI report for enhanced transparency.

These calls for regulation align with similar sentiments from OpenAI and Google. OpenAI’s CEO, Sam Altman, has testified before Congress advocating for the formation of a new regulatory entity and underscored the necessity of independent audits for AI models. Despite this, Altman has also hinted at the possibility of withdrawing OpenAI from the European Union if new policies remain unfavorable.

Kent Walker, President of Global Affairs for Google and Alphabet, echoed the need for collaborative efforts across government, academia, and industry to translate technological advancements into widespread societal benefits while addressing associated risks. The growing apprehensions regarding AI have even led some to step back from its development. Notably, Geoffrey Hinton, a former Google VP known as "The Godfather of AI," resigned in April to raise awareness about the potential dangers of AI.

In addition to expressing his concerns, Smith's speech highlighted the recent advancements in AI, likening its importance to that of the printing press or electricity. In a blog post complementing his remarks, Smith remarked that "AI offers perhaps even more potential for the good of humanity than any invention that has preceded it."