US dominance in AI needs to occur at every component layer, tech leaders say

Executives from prominent artificial intelligence technology companies told members of the Senate Commerce, Science and Transportation Committee Thursday that the global diffusion of the U.S. technology stack is critical for the U.S. to win the AI race.

Ensuring the U.S. produces and maintains control over the AI technology stack –– the individual components of an AI application, like its semiconductor chips, programming languages, interfaces and servers –– will determine if the country can lead the world in future AI innovation, witnesses told lawmakers. 

Lisa Su, the CEO of Advanced Micro Devices, said that the U.S. maintaining its competitive edge in AI innovation “actually requires excellence at every layer of the stack,” something that can be further enabled with “very supportive government policies.”

OpenAI CEO Sam Altman reiterated this posture, saying that the goal is for the entire world to be building its AI applications atop U.S.-built products. Altman made the case that global developers who don’t reference U.S.-made and -refined AI software tools for their individual software could create a digital threat.

“If someone's using a stack that we don't trust to train models, who knows what it's going to do? Who knows what back doors would be possible? Who knows what sort of data corruption issues could be possible?” he said. “I think the AI stack is increasingly going to be a jointly designed system, from the chip all the way up to the end consumer product, and lots of stuff in between.”

To help the world adopt as much of the U.S. stack as possible, Altman said OpenAI is releasing an open-source model this summer for the developer community to use in their software.

“Influence comes from people adopting U.S. products and services up and down the stack,” Altman said. “The most critical stuff, the creation of these models that will be so impactful that should still happen here.”

Altman and Su both commented on the importance of collaboration with a variety of experts in the public, private and academic sectors — a sentiment that Brad Smith, the vice chair and president at Microsoft, echoed in his testimony as well.

“If the United States is going to succeed in leading the world in AI, it requires infrastructure. It requires success at the platform level. It requires people who create applications,” Smith said. “Our success, each of our success, depends on each other's success.”

The tech community’s desire to evangelize all parts of the U.S.-made technology stack comes as the Trump administration is focusing on advancing the domestic U.S. supply chain and manufacturing through a new tariff regime and corporate partnerships. 

Other policy aspects the Trump administration is pursuing — such as light-touch regulation, domestic manufacturing, government adoption and energy generation investments — were all mentioned as vital to the U.S. winning the AI race. 

“This is an area where I think the devil is in the details, and it requires a lot of balance,” Su said. “And so from an industry standpoint, it's our job to put on the broader hat and work hand-in-hand with the administration and Congress to make our best recommendations.”