Opinion | Rise of ‘autonomous’ agentic AI raises the geopolitical stakes

Agentic AI will further level the techno-economic playing field and radically alter opinion-shaping and the military contest

Reading Time:3 minutes

Participants at the Responsible AI in the Military Domain (REAIM) summit in Seoul on September 10, 2024. Photo: AFP

Published: 8:30pm, 19 Feb 2025

Imagine a world where artificial intelligence (AI) agents could act as personal political advisers, negotiate for governments on foreign policy or carry out integrated military operations in war zones. These scenarios are increasingly possible with the emergence of agentic AI.

Unlike generative AI – large-language models (LLMs) that create content – the hallmark of agentic AI is its ability to make decisions and carry out complex actions autonomously, with minimal human supervision. Agentic AI usually consists of teams of AI agents, some with specialised roles – such as gathering information or carrying out tasks – while others act as team leaders or orchestrators of activities.

Agentic AI harnesses the cognitive abilities of LLMs, but combines these with other forms of machine learning and automation to make better decisions and take actions. Another key difference is agentic AI systems are goal-oriented, set to optimise particular objectives, such as maximising customer satisfaction or minimising the cost of a particular production process.

Although nascent, agentic AI is attracting attention from businesses and governments because of its potential to supercharge productivity and efficiency. What has received less attention is its potential to drastically change the dynamics of technopolitical rivalry between nations.

In particular, agentic AI is likely to alter the nature of the geopolitical contest across three interrelated spheres: techno-economics, influence and opinion-shaping, and the security, intelligence and military contest.

Until recently, the techno-economic sphere has largely been a race between the United States and China for AI supremacy (or at least parity for China). The electrifying entry of DeepSeek – a low-cost, largely open-source model – into the generative AI space in late January caused US technology stocks to tumble and shattered assumptions of American AI supremacy.