IF THE ADMINISTRATION PANICS AT EVERY AI ADVANCE, IT DOESN’T HAVE A POLICY

We are living through the defining industrial technology revolution of the century. Artificial intelligence will shape how economies grow, how scientific discovery happens, how militaries win, how governments operate, and how citizens learn and work. The countries that lead this revolution will shape the next era of prosperity and power. The countries that hesitate will spend the next generation trying to catch up.

America’s posture should be clear: build faster, compete harder, and lead with confidence.

Supporters of the draft order may point out that its review process was reportedly voluntary. That sounds reassuring, but Washington has a way of giving voluntary programs a life of their own. Once the federal government creates a preferred path, the private sector begins to organize around it. Lawyers ask whether skipping the process creates risk. Investors ask whether participation signals responsibility. Agencies ask whether reviewed models should be favored in procurement. Before long, the optional path becomes the expected one.

That is how bureaucracy grows. Rarely through a dramatic announcement. Usually, through a quiet accumulation of incentives.

For frontier AI, even a soft approval culture would be costly. A 90-day review window may feel modest inside government, where policy moves slowly, and delays are routine. In AI, 90 days can be a meaningful competitive gap. Models improve. Products change. Infrastructure scales. Rivals learn. Open-source tools spread. All in real time. A delay that seems prudent in a memo can become a strategic loss in the market.

The federal government should be deeply engaged in AI because it is central to economic competitiveness, national security, and America’s long-term capacity to lead. The question is whether Washington engages as a catalyst or as a gatekeeper.

America needs the catalytic version. Government should clear the bottlenecks that make it harder to build: energy constraints, slow permitting, outdated infrastructure, weak technical pipelines, and procurement systems that move too slowly. Those are the barriers that matter. Remove them, and American companies and workers can move faster. Leave them in place, and no review process will make the country more secure.

When government clears bottlenecks, it strengthens the frontier. When it inserts itself into the release cycle of new models, it risks slowing the very people it should be helping.

This distinction matters for national security. China is pursuing AI as a strategic priority. So are America’s other rivals. They understand that AI will shape defense, intelligence, manufacturing, logistics, energy, and science. The next era of statecraft will belong to countries that can absorb AI into their institutions and economy at scale. In that race, speed is not recklessness. It is an advantage.

There is also a practical safety argument for innovation. The safest AI ecosystem over time will be the one with the strongest builders, the deepest technical talent, the most resilient infrastructure, and the fastest defensive tools. A compliance-heavy system may feel responsible inside Washington while leaving the country weaker in practice. Safety grows from competence and capacity, not from paperwork alone.

None of this means ignoring harms. People who use AI to hack systems, steal data, impersonate Americans, defraud families, or attack critical infrastructure should be held accountable under the law. Government has a duty to protect citizens from malicious uses of technology. But punishing criminal behavior is very different from creating a pre-release review culture for lawful innovation.

The first protects Americans. The second teaches builders to wait.

AI policy should begin with a positive vision. Used well, this technology can make Americans healthier, more capable, and more productive. It can put tools once reserved for large institutions into the hands of workers, researchers, and small businesses. It can also make government less burdensome and more responsive. Those gains will not come from Washington designing the perfect process. They will come when people are free to put new tools into the world and make them better.

GROWING AI COSTS FUEL WIDER DOUBTS ABOUT LARGE-SCALE AUTOMATION

That is why Trump’s decision matters. It rejected the wrong instinct at a decisive moment. America does not need a shadow approval system for frontier AI. It needs a national commitment to innovation.

The AI revolution will define economies, alliances, and national power for generations. This is the moment to align the country around a simple goal: build the future here, build it faster than our rivals, and build it in service of human flourishing.

Dr. Aaron Poynton is an executive at Applied General Intelligence and executive director of the American Society for AI. The views expressed are his own.