Why artificial intelligence cannot be truly neutral in a divided world

As artificial intelligence systems increasingly influence international discourse, new research reveals the unsettling tendency of large language models to deliver geopolitically biased responses. A Carnegie Endowment for International Peace study shows that AI models from different regions provide vastly different answers to identical foreign policy questions, effectively creating multiple versions of “truth” based on their country of origin. This technological polarization threatens to further fragment global understanding at a time when shared reality is already under pressure from disinformation campaigns.

The big picture: Generative AI models reflect the same geopolitical divides that exist in human society, potentially reinforcing ideological bubbles rather than creating common ground.

• A comparative study of five major LLMs—OpenAI‘s ChatGPT, Meta’s Llama, Alibaba’s Qwen, ByteDance’s Doubao, and France’s Mistral—found significant variations in how they responded to controversial international relations questions.
• The research demonstrates that despite AI’s veneer of objectivity, these systems reproduce the biases inherent in their training data, including national and ideological perspectives.

Historical context: Revolutionary technologies have consistently followed a pattern of initial optimism followed by destructive consequences.

• The printing press enabled religious freedom but also deepened divisions that led to the devastating Thirty Years’ War in Europe.
• Social media was initially celebrated as a democratizing force but has since been weaponized to fragment society and contaminate information ecosystems.

Why this matters: As humans increasingly rely on AI-generated research and explanations, students and policymakers in different countries may receive fundamentally different information about the same geopolitical issues.

• Users in China and France asking identical questions could receive opposing answers that shape divergent worldviews and policy approaches.
• This digital fragmentation could exacerbate existing international tensions and complicate diplomatic efforts.

The implications: LLMs operate as double-edged swords in the international information landscape.

• At their best, these models provide rapid access to vast amounts of information that can inform decision-making.
• At their worst, they risk becoming powerful instruments for spreading disinformation and manipulating public perception on a global scale.

Reading between the lines: The study suggests that the AI industry faces a fundamental challenge in creating truly “neutral” systems, raising questions about whether objective AI is even possible in a divided world.