Is AI a Threat to Democracy? – The New AI Arms Race Explained

The New AI Arms Race: Beyond Geopolitics, a Battle for Influence in Everyday Life

The global competition surrounding artificial intelligence is often framed as a geopolitical struggle between the US and China, focusing on chip exports, model releases, and military applications. However, a more pervasive and arguably more significant arms race is unfolding across numerous sectors, with AI as the primary weapon and a diverse range of actors vying for advantage. This isn’t a contest for global supremacy, but a fragmented series of battles for influence within specific systems, and the stakes are surprisingly high.

From academic publishing to the legal system, open-source software development to journalism, and even the simple act of contacting elected officials, AI is disrupting established processes. Academic journals are grappling with a surge of AI-generated papers, prompting them to adopt AI-powered tools for review. In Brazil, the court system initially embraced AI to triage cases, but quickly found itself facing an influx of AI-assisted filings. Open source software developers are struggling to manage the volume of code contributions from automated bots. And the disruption extends to newspapers, music, social media, education, investigative journalism, hiring, and even government procurement.

Each of these represents an “arms race” – a dynamic where participants iteratively seek an edge by leveraging the same technology. The beneficiaries, however, are largely US mega-corporations, capturing wealth at an unprecedented rate. A significant portion of the global economy has shifted towards AI in recent years, a trend that is accelerating. Simultaneously, these companies’ lobbying efforts are increasingly shaping government policy.

Consider the evolving relationship between democratic governments and their citizens. Interactions that once occurred directly between people and their representatives are now being scaled up with AI taking on roles previously held by humans. The 2017 incident involving the US Federal Communications Commission’s net neutrality comment platform provides a stark example. The platform was flooded with millions of fraudulent comments orchestrated by broadband providers to oppose regulation, and countered by a single individual submitting millions of pro-regulation comments using automated tools. While the tools used then were rudimentary, the principle remains the same.

Today, it’s becoming increasingly difficult for citizens to discern whether they are interacting with a government bot or if online discussions about policy are simply bots communicating with bots. In the US Congress, staff are already utilizing AI to manage constituent email correspondence. Politicians campaigning for office are employing AI tools for fundraising and voter outreach. Estimates from 2025 suggest that as much as a fifth of public submissions to the Consumer Financial Protection Bureau were already AI-generated.

This shift is driven by a fundamental problem: traditional mass advocacy campaigns have often been ineffective due to the inverse relationship between quantity, quality, and relevance. Government agencies often prioritize specific, actionable comments over generic submissions. AI addresses this by enabling easy contextualization and personalization, allowing individuals to make their voices heard more effectively. As the volume and detail of constituent comments increase, agencies inevitably turn to AI to manage the influx.

This creates a cyclical arms race. Increased AI-driven submissions necessitate AI-powered analysis on the receiving end. While any temporary advantage gained by one side is likely fleeting, real harm is created when one party exploits the system. Democracies suffer when public servants use AI-generated responses to dismiss constituent concerns, and the scientific enterprise is undermined when fraudulent AI-generated papers overwhelm legitimate research.

As outlined in the new book, Rewiring Democracy, this arms race dynamic is inevitable. Every actor in an adversarial system is incentivized to leverage new technologies to advance their interests. However, opportunities exist to harness AI for positive change. The key is to both utilize the technology to one’s advantage and resist the concentration of power it enables.

Currently, the primary beneficiaries of this AI boom are a handful of American Big Tech corporations, extracting trillions of dollars from the development of AI chips, data centers, and “frontier” AI models. Regardless of which side prevails in each arms race scenario, these corporations consistently profit from the dynamic itself. However, democracies are actively resisting this concentration of wealth and power through anti-trust regulation, human rights protections, and the development of public alternatives to corporate AI.