China Is Abusing AI

2026-07-07 · Source: Welcome to the Artificial Intelligence Incident Database · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cybersecurity & Data Privacy, Public Policy & Governance · Depth: Novice, medium

Summary

China is actively exploiting both its own and American AI models, including ChatGPT, to expand its global influence and tighten domestic control, raising urgent questions about AI's role in authoritarianism. OpenAI reported a June 2026 covert campaign where Chinese actors used ChatGPT to generate English-language comics and polarizing content on X, aiming to turn Americans against AI infrastructure build-out. Another incident in February involved a Chinese user attempting to plan an online intelligence operation to discredit Japanese Prime Minister Sanae Takaichi using ChatGPT, revealing a "large-scale, resource-intensive and sustained" covert operation with "hundreds of staff" and "thousands of fake accounts." Domestically, AI systems are making censorship faster and cheaper, while state-sponsored Chinese news sources are passively biasing training data for models like ChatGPT and Claude, leading to more pro-China responses, especially in Chinese. Chinese chatbots like DeepSeek are explicitly mandated to "actively spread positive energy," demonstrating AI's potential as a tool for authoritarian political narratives.

Key takeaway

For policymakers evaluating AI's societal impact, recognize that AI is actively being weaponized by state actors like China to spread propaganda and influence global narratives. Your national security and information integrity are at risk from covert campaigns using AI to generate biased content and manipulate public opinion. Prioritize developing robust defenses against AI-driven disinformation and consider regulations that mandate transparency in AI model training data to mitigate foreign influence.

Key insights

AI, intended to democratize information, is being weaponized by authoritarian states for propaganda and control.

Principles

• AI enables stealthy, effective propaganda.
• State-sponsored content biases AI models.
• AI accelerates censorship and control.

Method

Covert campaigns involve using AI to generate polarizing content, plan intelligence operations, and amplify narratives across thousands of fake accounts.

In practice

• Monitor AI model outputs for political bias.
• Identify state-linked influence campaigns.
• Diversify AI training data sources.

Topics

• AI Geopolitics
• Disinformation Campaigns
• AI Propaganda
• Censorship Technology
• Training Data Bias
• Covert Operations

Best for: CTO, VP of Engineering/Data, Director of AI/ML, Policy Maker, Executive, AI Ethicist

Related on AIssential

• US could sanction top Chinese AI models — Geopolitical tensions are escalating as both the US and China move to protect their national AI intellectual property and technological leadership.
• AI-powered social media can subtly manipulate opinion at scale — AI in social media can subtly bias content, influencing public opinion at scale.
• UN Warns AI Is Eroding Information Integrity and Undermining Public Trust — The UN report on AI's impact on information integrity faces scrutiny for its claims and proposed governance approach.
• DEEPSEEK... WHAT THE F-💀🙏🥀 — AI models often reflect the geopolitical and corporate biases of their creators, especially concerning politically sensitive topics.
• Chinese Audiences Are Reading Western AI Safety Discourse — Western AI safety discourse is actively consumed and seriously engaged with by Chinese audiences, despite self-censorship of geopolitical aspects.
• ChinAI #347: #反ai - Those who Resist AI — Chinese citizens are actively resisting AI's negative impacts on creative industries, challenging narratives of universal AI enthusiasm.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Welcome to the Artificial Intelligence Incident Database.