5 June 2025

Selective Openness: America's Path to Technological Dominance

Jags Kandasamy

As Congress debates the American Innovation and R&D Act of 2025, the United States is facing an existential test: how to maintain its technological supremacy in a world marked by increasingly intense global rivalries, especially with China. This isn’t just a competition of money; it's about national security. China’s version of America’s GPT-4 model, the DeepSeek AI model, has already shrunk America’s lead in artificial intelligence to months and underscored how our defense innovation ecosystem is still broken.

The act is directed at STEM education, semiconductor manufacture, and secure supply chains. It is a significant plus-up to help us retain U.S. leadership in AI, quantum computing, and 5G — the technologies underpinning our future military capabilities. Lawmakers must act urgently to balance open innovation and strategic defense, or risk watching adversaries exploit that balance on the battlefield and beyond.

Without question, America has maintained technological dominance for decades by sharing technology among open-source contributors. From Silicon Valley start-ups to America’s most prominent defense companies, the Pentagon has emerged as arguably the driver of a staggering array of technologies, like artificial intelligence and cloud computing, now promising to transform how we live and fight. But untrammeled transparency today threatens to destroy the edge we still have. China's deployment of open-source AI models powered by its great digital wall, such as the models DeepSeek relies on, is part of an ongoing process refining the country's AI military, including systems for autonomous drones and cyber warfare and at the same time, denying Americans and the rest of the world access to its ingenuity.

We offer a more nuanced and balanced solution to this dilemma—one that carefully demarcates the technologies that should remain open to international collaboration and those critical innovations we must guard more zealously.

First, broad openness ought to prevail for technologies that create vast opportunities for economic growth and positive societal application but carry little military risk. Examples like healthcare AI, general-purpose machine learning frameworks, and technologies aimed at the public good should be collaborative, global efforts.

No comments: