Artificial intelligence technology. Image: Wikimedia Commons
Artificial intelligence technology. Image: Wikimedia Commons

China wants to turn artificial intelligence into an offensive battlefield tool, but Chinese military researchers say the People’s Liberation Army (PLA) still has a long way to go.

A recent study found that the PLA mainly uses AI for support tasks such as answering questions and generating content, while its potential combat applications remain largely exploratory, the South China Morning Post reported.

The researchers further noted that the country’s AI models “still lag considerably behind foreign large language models in performance.”

To close the gap, they proposed building an “autonomous and controllable AI attack chain” that would move the technology beyond support functions and into offensive operations.

From Assistant to Attacker

The proposed attack chain would use AI to identify and engage battlefield targets, generate code for attacks on enemy networks and command systems, and influence adversaries through cognitive warfare.

The study draws heavily on US military research into AI-enabled cyber and information warfare, including the technology’s potential to support attacks, disrupt physical infrastructure, and shape battlefield decisions.

However, weaker models are not the PLA’s only obstacle.

The researchers also identified a shortage of military personnel with expertise in AI and big data, noting that defense adoption has struggled to keep pace with China’s rapidly advancing civilian technology sector.

Recruits from private industry may bring strong technical skills but often lack military experience, potentially limiting their effectiveness in operational environments.

You May Also Like

Defense Cyber Ops Get Agentic AI Boost as US Startup Raises $17.5M

The startup’s Caspian platform aims to unify fragmented cyber defenses in sovereign and contested environments.

Texas AI Startup to Fast-Track Pentagon’s Digital Overhaul

The project will integrated solutions in support of the Pentagon’s cybersecurity, machine algorithm modeling, applied data science, autonomous systems, and DevSecOps.

AI to Help US Air Force Defend Network of 700,000 Users

Crogl’s technology will work across the Air Force’s existing security infrastructure, automating much of the threat investigation while keeping response decisions with human analysts.