The CHIPS and Science Act, signed in August 2022, set aside $52.7 billion for US semiconductors, including $39 billion in manufacturing incentives run by the Commerce Department’s CHIPS Program Office. Those incentives pay for new fabs, packaging plants and suppliers of wafers, gases and equipment, mostly as direct grants, sometimes with loans. Most of the big awards were finalized in the last months of the Biden administration. The Trump administration kept the programme but changed how it works, most visibly by turning Intel’s grants into a government shareholding.

Most laws aimed at the most powerful AI models decide who is covered by counting training compute. The unit is FLOP, the total number of floating-point operations used to train a model. This page compares the thresholds now written into law, what each one triggers, and when the obligations start. Last updated: 15 September 2026. The short version The EU sets its line at 10^25 FLOP. California, New York, Illinois and South Korea all use 10^26, which is ten times higher.

AI sovereignty is the capacity of a state to develop, deploy and govern artificial intelligence systems without depending on infrastructure, models or rules controlled by another state. The term entered mainstream policy vocabulary around 2023 and has since been applied to national compute programmes, domestic model development, data residency law and regulatory regimes. It is used loosely, and that looseness matters: the phrase bundles together at least four distinct claims with radically different price tags and success rates.

The same carbon fiber that stiffens a racing bicycle frame can reinforce a ballistic missile’s nose cone. The nickel superalloys machined into jet turbine blades are equally at home in the combustion chambers of cruise missiles. Dual-use materials — substances whose physical properties make them valuable in both civilian industry and military hardware — sit at the intersection of commerce, science, and national security in ways that no clean regulatory line has ever fully resolved.

The federal government’s push to consolidate invention reporting onto a single platform is the right instinct—but the platform has structural limitations that are creating new inefficiencies even as it eliminates old ones. A July 2023 executive order directed nine major agencies to transition to iEdison, NIST’s web-based invention reporting system, by the end of 2025. GAO’s April 2026 technology transfer report (GAO-26-107971) finds the transition unfinished and the system itself in need of improvement.

Federal agencies fund billions of dollars in research every year. Some of that research produces inventions. Under a 1980 law called the Bayh-Dole Act, the universities, small businesses, and nonprofits that receive this federal funding can keep ownership of the resulting inventions—provided they meet certain reporting requirements. The theory is straightforward: give researchers and institutions a financial stake in commercializing their discoveries, and more federally funded innovation will reach the public as useful products.

When a university or company invents something using federal research funding, the Bayh-Dole Act gives them the right to keep it. So why do about one in five choose not to? GAO’s April 2026 technology transfer report (GAO-26-107971) provides the most detailed public accounting of this question to date, drawing on five years of invention disclosure data from 30 federal agencies covering fiscal years 2020 through 2024. The data reveals a practical picture of how institutions make IP decisions—one that has more to do with commercial realism than with any failure of the incentive structure.