Photo credit: Bolivia Inteligente
Google has released a paper outlining its proposed two-pronged approach to AI governance in the US, doubling down on “fair use” AI training.
Late last week, Google released a policy paper outlining a proposed approach to AI regulation in the United States that would separate oversight of widely deployed AI applications from pioneering AI models. To that end, the company is calling for a “pragmatic, evidence-based framework” that treats advanced systems differently than tools like chatbots.
Unsurprisingly, the paper argues that AI policy should focus on the outputs of a particular model, rather than the inputs used to create it. Google is proposing the creation of the Frontier AI Regulatory Organization (FARO), an industry-funded organization that would operate under federal oversight and develop standards for security, safety, transparency, and incident reporting.
Google says FARO has the potential to set scientific benchmarks for the capabilities of frontier models in areas such as chemical, biological, radiological, and nuclear risks, as well as cybersecurity. FARO would also oversee independent audits and potentially require cutting-edge AI companies to publish and follow safety frameworks before releasing advanced models.
Google, on the other hand, argues that the federal government should primarily rely on existing legal frameworks when dealing with widely deployed AI applications, with targeted updates as needed. Google says policy should focus on “real-world outcomes and specific harms” rather than “micromanaging the science” behind these tools.
“We believe a balanced approach can protect the public without paralyzing progress and ensure everyone can benefit from AI,” said Kent Walker, Google’s president of global affairs. “We hope these ideas will contribute to the important conversations that are ongoing, and we encourage you to read the full text.”
Google has cited several priority areas, including child safety, information integrity, workforce readiness, copyright, privacy, and data center energy infrastructure. The company also supports measures such as AI interaction guidelines for children, disclosure that chatbots are not sentient, rules for queries related to self-harm, watermarking and provenance standards for generated AI, employee upskilling, and privacy-enhancing technology.
