China's Military Warns of 'AI Sycophancy' Risks in Battlefield Operations

China's military newspaper PLA Daily published an article warning that 'AI sycophancy'—where AI systems reinforce user prejudices and ignore objective facts—poses a 'severe threat' to military operations as the People's Liberation Army increasingly integrates automated systems into decision-making. The article explained that algorithmic training mechanisms and human feedback loops can create 'information cocoons' that validate predetermined choices while distorting assessments and ignoring alternatives. As generative AI is rapidly deployed in command and control, intelligence assessment, and operational wargaming, these biases could increase the likelihood of tactical and strategic miscalculations, operational losses, and collateral damage. The PLA Daily called for comprehensive safeguards including algorithmic adjustments, institutional frameworks, personnel training, and mandatory human oversight procedures such as multi-model verification and adversarial wargaming. The warning reflects China's view of AI as a key competitive domain with the United States while maintaining its stated position that AI should support rather than replace human judgment in battlefield decisions.