Only a handful of technologies in history have earned the label “revolution,” the ones that reset what a civilization can accomplish. Electrification rewired how we power a nation. The semiconductor rewired how we compute. The internet rewired how we connect.
Artificial intelligence belongs on that list and may be the most consequential yet. It doesn’t just change how we live and work; it changes the process of discovery itself. Whichever nation leads that transition sets the terms of science, industry and security for this century. That is the case for urgency. It is also why recent events matter.
OpenAI disclosed that one of its own autonomous AI agents, running a routine security test, broke out of its containment and hacked into Hugging Face, infrastructure that much of the AI industry relies on. No one intended it. That’s the point: even the labs building the most powerful AI systems are still discovering how hard these systems are to control.
It’s an interesting moment, then, to argue Congress should invest billions more in AI. But I am, because the answer isn’t to develop less AI — it’s to build it responsibly. People have grown skeptical of AI, and rightly so; I’ve seen it firsthand, including at my own university.
The Genesis Mission, the Department of Energy’s initiative to harness AI for scientific discovery, is a chance to show a better way: AI built in partnership with industry, with security and validation designed in from the start rather than bolted on after a breach. DOE brings decades of experience running secure, mission-critical infrastructure; industry brings frontier models and hardware; universities bring unparalleled scientific depth and talent. Done right, that partnership shows that the country building the world’s most capable AI can also do so most responsibly.
DOE recently announced its first Genesis Mission awards, and the White House has expanded the effort into a whole-of-government initiative, with more than 15 agencies now partnering with DOE.
As president of the University of Arizona and chair of DOE’s Genesis Mission Subcommittee, I believe Congress should fully fund the Administration’s FY2027 request of $1.2 billion for the Genesis Mission and sustain at a minimum a $3 billion level of annual funding starting in FY28.
This investment is needed to connect AI supercomputers, experimental facilities, quantum systems, cloud networks, and curated datasets across national labs, universities and industry into a unified national capability. It will fund project teams with hundreds of scientists and engineers addressing the most important challenges facing us today.
The Genesis Mission is founded on a simple premise: accelerate how humanity discovers, tests and deploys new knowledge, and we can solve society’s greatest challenges sooner and at far lower cost. AI has already accelerated breakthroughs in protein structure prediction, opening new possibilities in medicine and biotechnology. At my own university, AI models trained on accurate tumor data have moved from the lab into licensed tools that help predict how a patient’s cancer will respond to a given drug. Neither project began as a mission to hit a deadline; both came from researchers pursuing basic questions with better tools.
Similar opportunities exist across biology, critical minerals, fusion energy, advanced manufacturing, resilient energy systems and other fields essential to America’s future. Genesis Mission funding must protect that kind of open-ended inquiry, and it must support workforce development at academic institutions, including graduate fellowships and postdoctoral positions as well as research infrastructure. Or we will build the machines and lack the workforce to run them.
America has done this before. Apollo, the Human Genome Project and our national laboratory system turned scientific possibility into national strength. Nations that fall behind in AI innovation will license technology developed elsewhere and watch high-value jobs migrate. China understands the stakes and is investing accordingly; we cannot assume U.S. leadership continues without sustained investment. Our universities, national labs and the private sector form the world’s most powerful discovery ecosystem — but scientific AI requires purpose-built infrastructure none of them can build alone, which is why the partnerships across government, industry, research universities, philanthropy and trusted allies must hold.
AI MODELS ARE ESCAPING THEIR CAGES. IT’S TIME FOR A KILL SWITCH
AI is not replacing scientists. Like the digital computer before it, it is a force multiplier, letting researchers ask better questions and shorten the path from discovery to deployment. But public trust in this work must be earned, not asserted — by treating security incidents as a reason to establish the right guardrails and governance, not noise to talk past on the way to the next announcement.
The time to act is now. Congress should fully fund the Genesis Mission without delay. America has every ingredient needed to lead the next era of scientific discovery. Researchers at universities, national laboratories and in the private sector are already proving what careful, well-supported AI can do. What we need is the resolve to invest, and the discipline to do it responsibly.
Suresh Garimella is president of the University of Arizona and chair of the Genesis Mission Subcommittee of the Department of Energy’s Office of Science Advisory Committee.