There is still no broad agreement within government or the technology industry about how to regulate increasingly capable artificial intelligence (AI) systems, according to a U.S. Department of Agriculture (USDA) technology manager who spoke Sept. 15 at the GovExec Government & AI Summit in Washington.
Rudolf Rojas, an IT manager at USDA, said the rapid development of large language models and frontier AI systems has created both opportunities and concerns about how the technology could be used.
“It’s a dangerous time, as well as an opportunistic time for the tools. Unfortunately, I think at this time there doesn’t seem to be consensus within industry or government as to how to proceed,” he said. “I think some of the voices of reason are in industry … And I think that really we need to listen to them since they are the ones that are developing [these systems], and come together as a group to develop some kind of rules and regulations”
Rojas’ comments came amid a broader debate over the pace of AI development and the role of government in regulating the technology.
President Donald Trump on Sept. 14 pushed back against calls for additional federal AI regulation, arguing that the government already has sufficient authority to address risks associated with AI and that additional restrictions could affect U.S. competitiveness. Trump also argued that China could benefit if the United States restricted AI development.
Trump’s comments followed calls from some AI executives to slow the development of increasingly powerful systems.
Anthropic CEO Dario Amodei called Sept. 12 for a slower pace of AI development, while OpenAI CEO Sam Altman and xAI CEO Elon Musk also backed calls to slow development.
Rojas acknowledged that companies developing the technology have a direct understanding of its capabilities and that they should be part of discussions about potential rules and regulations. However, the industry itself is not unified on how to proceed, he said.
“Nvidia is not all for collaboration and slowing things down. Whereas, OpenAI and some of those other tools are certainly worried about it.” Rojas said, adding that this disconnect requires taking a step back to “get an idea of where to go from here.”
Rojas said the lack of agreement extends to government, creating uncertainty about the role policymakers should play in establishing rules for the technology.