During my time on the House intelligence committee, one thing became abundantly clear: America’s adversaries never stop looking for new ways to exploit our vulnerabilities.
The threats evolve, the technologies change, but our mission remains the same: anticipate threats before they reach our shores and ensure the United States maintains the strategic advantage.
Recommended Stories
Today, that mission is more urgent than ever as artificial intelligence changes the speed and scale of the threats we face. In the right hands, AI can help America identify weaknesses, harden our defenses, and stay ahead of our adversaries. In the wrong hands, it can help those adversaries find and exploit vulnerabilities faster than ever before.
America’s closest intelligence partners recently sounded the alarm.
The Five Eyes alliance — the U.S., the United Kingdom, Canada, Australia, and New Zealand — together warned that artificial intelligence is rapidly transforming cyber risk, accelerating the speed, scale, and sophistication of threats facing government, businesses, and critical infrastructure.
Our adversaries understand exactly what is at stake.
China has invested enormous resources in artificial intelligence because Beijing recognizes that dominance in AI will shape not only the global economy but also the balance of military and intelligence power for decades to come.
Russia, Iran, North Korea, and others are pursuing many of the same capabilities.
The question facing the U.S. is not whether AI will change the cyber battlefield. It already has. The question now is whether America will ensure that those responsible for defending our most essential systems have access to the tools necessary to stay ahead.
President Donald Trump’s recent executive order on AI innovation and security shows that this administration understands a critical truth: America’s leadership in AI is inseparable from our leadership in national security.
The order takes an important step by directing the Department of Homeland Security’s Cybersecurity and Infrastructure Security Agency to help agencies, state and local authorities, and critical infrastructure operators gain access to advanced AI cybersecurity tools.
It also creates a voluntary framework for advanced AI developers to work with the federal government to select trusted partners for early access before those models are released to the public.
That is the right foundation. Now the administration should build on it by ensuring that trusted critical infrastructure operators are included in that early-access framework as well.
Federal agencies cannot defend the nation alone. They rely every day on privately operated networks, systems, and services. If those systems are vulnerable, government operations are vulnerable too. Telecommunications companies and broadband providers, for example, don’t just serve millions of individual consumers — they also underpin military readiness, emergency response, healthcare delivery, financial systems, education, and nearly every aspect of modern life.
Our adversaries understand this reality. They continue to target the networks Americans use to communicate, do business, call for help, and access essential services. They know that if they can compromise those networks, they can disrupt far more than phone calls and internet service. They can put pressure on the systems that keep the country running.
That is why trusted critical infrastructure operators need access to the best AI tools to play defense well before malicious actors can access those same tools to play offense. Banks, hospitals, electric utilities, telecommunications providers, and broadband networks should not be left waiting while hostile actors learn how to use the next generation of AI against them.
The administration should build on its early leadership by establishing a clear, secure process for these trusted operators to receive early access to frontier AI models before broad public release. That process should be thoughtful and deliberate; it should require secure testing, protect sensitive information, and keep CISA closely involved.
WHY TRUMP V. SLAUGHTER WON’T SCARE OFF AI INVESTMENT
The goal is simple: Find the weaknesses before our enemies do.
Trump’s order points the country in the right direction. The next step is to make sure the operators who defend the systems Americans depend on every day have the tools they need to stay ahead.
Christopher Douglas Stewart is an American politician, author, and businessman who served as the U.S. representative for Utah’s 2nd District from 2013 until his resignation in 2023.