Trump launches federal initiative to accelerate AI-driven scientific discovery

On November 24, President Trump signed an executive order creating the Genesis Mission, a sweeping federal effort to pull together the country’s most powerful scientific tools—AI, supercomputing, and decades of government-funded research data—and put them to work in a more coordinated way. The goal is straightforward but ambitious: speed up breakthroughs in medicine, clean energy, and national security by making it easier for researchers to tackle scientific questions that currently take years of modeling or lab work.

At the center of this initiative is the Department of Energy, which already oversees many of the country’s high-performance computing centers and national laboratories. Genesis asks DOE to knit these resources into a unified AI-enabled system. That includes building a platform for generating scientific foundation models, running massive simulations, and even developing robotic labs that can test hypotheses with minimal hands-on work. Oversight will run through the White House Office of Science and Technology Policy, along with a newly appointed Special Advisor for AI and Crypto, plus academic and industry partners.

White House officials say they want this platform to double as a safe testing ground for new AI systems, somewhere researchers can evaluate emerging models before they’re integrated into high-stakes scientific work. If it works as intended, the structure could cut down on redundant efforts across agencies, speed up review cycles, and ensure that different labs are using more consistent technical foundations.

A scale of U.S. science mobilization not seen in decades

Analysts at the American Institute of Physics noted that the scope resembles the sweeping federal reorganizations behind the Manhattan Project and the Apollo program, two efforts that reoriented national scientific capacity toward urgent priorities.

In many ways, Genesis is the next chapter in a much longer story. Federal agencies have spent over half a century investing in AI and advanced computing. DARPA and NSF seeded early machine learning research in the 1960s. DOE labs later pushed the frontier on supercomputing, moving from early Cray machines to the exascale systems now used for climate modeling, nuclear science, and materials research.

Those labs also sit on enormous amounts of scientific data, much of it trapped in outdated formats or living inside specialized research environments. Genesis reimagines those archives as training fuel for modern AI models, and calls for a government-wide effort to clean, standardize, and prepare decades of research data for today’s multimodal workflows.

Where Trump aims to shift scientific power

The administration emphasized fields where scientific competition has global stakes. Biotechnology has become central to drug development and pandemic readiness. Semiconductor research underpins national supply chain strategy and the CHIPS and Science Act of 2022. Quantum information science is viewed as essential for secure communications. Clean energy programs, including fusion and fission research, now anchor both climate and energy-independence planning.

Supporters say scientific foundation models could accelerate progress across these areas. A 2023 workshop from the National Academies of Sciences described how AI systems trained on multimodal data could help identify new materials, optimize chemical processes, and surface therapeutic targets in far less time than traditional laboratory cycles. The report emphasized the promise of closed loop systems in which simulation, experimentation, and analysis inform one another continuously.

The Genesis Mission builds on work already underway at large research institutions. The National Institutes of Health has spent years preparing biomedical datasets for AI and establishing standards for transparency, equity, and reproducibility.

But researchers caution that the technical path ahead is difficult. A 2025 interdisciplinary analysis published on arXiv argued that large scientific models depend on carefully curated data, sustained collaboration between domain scientists and machine learning experts, and robust evaluation frameworks to confirm that AI generated hypotheses translate to real world results. The authors warned that automated discovery can produce misleading findings if evaluation systems are not rigorous.

Data quality remains a core concern. Federal archives contain decades of research, much of it created before modern metadata practices. Preparing these data for AI training will require substantial investment, similar to modernization efforts seen in health care and climate science.

Can Genesis deliver?

The mission also reflects rising global competition over AI-driven research. Analysts have pointed to parallels with China’s national strategies, which link supercomputing capacity with automated laboratories. Reuters reported that the administration views Genesis as essential to maintaining U.S. leadership in emerging technologies and reducing dependence on foreign supply chains.

The next phase will hinge on execution. DOE must design the experimentation platform, prepare datasets for AI workflows, and coordinate across laboratories with different missions. Universities and private companies will need clear mechanisms for contributing data and accessing federal infrastructure. OSTP is expected to issue guidance that covers research security and intellectual property, along with standards for publication and industry transition.

If the Genesis Mission succeeds, it could reshape the pace of American science by bringing supercomputing and AI-driven experimentation into a single national ecosystem. Supporters believe the approach could accelerate work in drug discovery and clean energy, as well as advanced materials research. If it falls short, the country may be left with powerful new tools, but without the governance or integration needed to use them effectively.