Accelerating Innovation, Performance, and Impact at USAID
Summary
The United States Agency for International Development (USAID) tackles some of the toughest challenges in some of the toughest places on earth, including fighting global pandemics, growing economic prosperity, strengthening democratic institutions, and providing humanitarian relief. USAID plays an important role in addressing global development problems that initially lack effective, scalable, and sustainable solutions. Yet USAID’s activities allow little room for the risk taking or iteration needed to drive significant improvements and encourage disruptive new ideas, with most programs implemented through detailed plans with rigid designs.
Imagine if the pace of progress for global development could match the breakneck pace of advances in the tech sector. The next administration should assess where current USAID interventions are inadequately meeting global need, applying best practices from innovation to improve programs accordingly. This will include shifting to outcomes-based performance metrics, dedicating budget for experimentation, establishing incentives that encourage risk-taking, linking payments with outcomes, and conducting ex-post evaluations of scale and sustainability.
At this inflection point, the choice is not between speed and safety but between ungoverned acceleration and a calculated momentum that allows our strategic AI advantage to be both sustained and secured.
Improved detection could strengthen deterrence, but only if accompanying hazards—automation bias, model hallucinations, exploitable software vulnerabilities, and the risk of eroding assured second‑strike capability—are well managed.
A dedicated and properly resourced national entity is essential for supporting the development of safe, secure, and trustworthy AI to drive widespread adoption, by providing sustained, independent technical assessments and emergency coordination.
Congress should establish a new grant program, coordinated by the Cybersecurity and Infrastructure Security Agency, to assist state and local governments in addressing AI challenges.