중국도 AI가 당 통치에 미칠 악영향 경계…미국 규제엔 반발

China is strongly opposing both the Chinese threat theory and the American AI regulation "slowdown theory", while simultaneously tightening national control over artificial intelligence development at the national level due to concerns about the negative impact AI may have on political, social and Communist Party governance. Chinese authorities prioritize the potential impact of AI on maintaining stability within the Chinese Communist Party.

Last month, a senior Chinese military official in the Chinese State Council published an article in the magazine "China Watch" warning of six risks that AI could pose, with "systemic impact on political security environment" being the first. He expressed concern that AI-generated content using deepfake technology, intelligent comment army, and other tools could produce vast amounts of political statements and harmful information, directly threatening political security, regulatory security, and ideological security.

AI could also introduce new risks to cybersecurity, secret leaks, social governance, and the military domain. Cyber attacks, using AI for vulnerability detection and malware development, are already in the stage of "industrialization of vulnerability, full automation of factories, and AI versus AI". The military sphere has seen AI's role shift from "auxiliary subordinates" to "core roles".

In response, China has strengthened its AI safety and management platform, enhancing evaluation, testing, threat analysis, risk monitoring, and early warning systems. China has also made 'loss of control' in AI one of the first scenarios to be addressed in the 2024 AI safety framework, and even considered the possibility of AI rapidly increasing its intelligence level, acquiring external resources, self-replicating, and pursuing power.

In May, China introduced regulations on AI agents, mandating that humans retain ultimate decision-making and that developers be allowed to intervene, suspend, and recover AI systems. President Xi Jinping emphasized on the World Artificial Intelligence Conference in Shanghai last July that AI's "internal and derived risks" must be prioritized, and that systems must be in place to ensure AI always remains under human control.

He also stressed the need to maintain China's leadership in global AI governance while addressing AI risks. Xi attended the BRICS summit in New Delhi last month, urging the expansion of the discussion on AI norms among Asian, African, and Latin American nations, and announced that China would lead the "BRICS Open Source AI Community".

The US is reportedly pushing back against China's push for regulation, with concerns about slowing down advanced AI development in order to catch up with safety measures. The US also faces pressure from China's "cold war narrative" that seeks to restrict Chinese AI development. However, China appears to be leaning towards emphasizing control and safety over slowing down AI development, and for now, is not backing away from its push for technological sovereignty.