OpenAI and the White House have competing visions for regulating artificial intelligence

Source Domain: mashable.com

In a recently released policy paper entitled “Democratic Governance of Frontier AI: A blueprint for a federal framework,” OpenAI put forward its vision of AI regulation, built around five core priorities: promoting transparency, protecting innovation, addressing risks to national security and public safety, advancing democratic governance, and creating “adaptive institutions” capable of keeping up with these rapid technological developments.

But while those are all laudable goals, there is very little agreement on how to pursue them in practice. And according to reporting by Politico, the timing of this paper is auspicious, coming shortly after the White House released two executive orders on “Promoting Advanced Artificial Intelligence Innovation and Security” that would place AI regulation squarely within the government’s remit.

As Politico AI reporter Brendan Bordelon points out, the OpenAI paper is an attempt to “nudge” the federal government towards a different approach, one in which civilian institutions are responsible for AI oversight. Outlining a process they call “reverse federalism,” OpenAI proposes that states be allowed to “to develop and refine common legal frameworks first,” before Congress adopts them at the national level.

Mashable Light Speed

In their vision, the Center for AI Standards and Innovation (CAISI) would act as the main point of contact between artificial intelligence companies and the government, working off a precedent set earlier this year when Microsoft, Google, Anthropic, and OpenAI all signed deals with the Commerce Department to allow the federal government to evaluate their AI models for potential national security risks.

As AI becomes more ubiquitous across the country, regulators are struggling to catch up. Between the potential for massive job losses, even in manufacturing, and the turmoil caused by AI deepfakes…

Source