In October 2019, Brookings and Tsinghua University organized a dialogue on AI and national security between the US and China. AI development leaders must prevent military use that harms civilians through governance and preparedness regimes. Governments can implement agreements to reduce the probability and consequences of AI-driven military crises and regulate their technical and use causes.
AI is a diverse field of interconnected techniques, used in applications such as molecular property prediction, video enhancement, and multilingual voice generation. This influences strategies and regulations, requiring specific testing and risk assessment. A distinction is made between narrow and general AI, with specialized models for specific tasks and others trained on broad datasets. The diversity of models reflects the evolution of human intelligence in domains such as visual perception, agent control, and 3D pose estimation.
The dominant countries in AI, such as the United States and China, generate geopolitical and economic tensions. The European Union seeks to develop its own sovereign AI to reduce its technological dependence and foster innovation based on democratic values.
In 2025, 47 global states introduced artificial intelligence legislation, with democrats focused on citizen protection and republicans on banning harmful uses and promoting innovation. Key issues include intimate image protection, elections, transparency, automated decisions, government use, employment, and health.
The Trump and Biden administrations have agreed on five key points regarding the regulation of artificial intelligence (AI), recognizing the need to regulate its use and acquisition by the federal government. The points include the identification of high-impact AI systems, the creation of governance and reporting structures, the prohibition of use without active protections, minimum practices for these systems, and transparency in their use.
