AI and the digital divide

Generative AI models such as GPT-4, Qwen2.5-Max, Deepseek v3, LLama, and Gemini 2.5 are trained on massive datasets sourced from publicly available information, text, images, and code. Most foundational models have been developed by leading tech companies like OpenAI, Google, Anthropic, Meta, Alibaba, etc. The AI companies control product access through subscription models, licensing fees, or pay-per-use pricing structures. Premium versions of generative AI tools often come with higher performance levels, faster response times, or exclusive features unavailable in free tiers. This commercialization creates barriers to access for individuals, small businesses, researchers, and even governments in low-income regions who cannot afford the costs associated with these services or don’t have the proper infrastructure and facilities to use AI.

AI models personalize learning experiences, diagnose diseases earlier, optimize logistics, fight climate change, and enhance security. As AI becomes essential to modern infrastructure, comparable to electricity grids or the internet, its privatization and concentration in the hands of a few companies raises concerns about accountability and transparency. According to the World Economic Forum, 2.5 billion people still don’t have access to the internet, and over 30 percent of the global population can’t take advantage of internet services. Moreover, 40 percent of international investments in research and development are made by just 100 companies. AI inequity is similar to many problems we face regarding wealth distribution, internet access, and access to education. Some argue that the global community is morally obligated to ensure that developing economies are not left behind in the AI revolution. These advocates call for more public policy, international cooperation, and potentially open-access AI models to reduce the digital divide. Opponents argue that a competitive free market remains the most efficient and effective way to drive innovation. From their perspective, market-driven models will benefit everyone as technological advancements eventually become more affordable and widely distributed.

The graph below shows the market capitalization of the five largest companies. Their combined value exceeds the GDP of the entire African continent and accounts for over ten percent of the global GDP.

In this debate, you either favor government intervention or the free market. You may argue for regulation, subsidies, public research funding, open-source alternatives, or international cooperation to ensure AI benefits are shared globally. If you favor the free market, you may argue that minimal interference will eventually promote competition, efficiency, innovation, and global scalability. If you look at the model of the USA, you could argue that the free market made the USA dominant in AI in the first place.

Resources

Ben-Avie, J. (2024). Don’t let AI become the newest digital divide. Council on Foreign Relations.

https://www.cfr.org/blog/dont-let-ai-become-newest-digital-divide

International Labour Organization. (2024). Mind the AI divide: Shaping a global perspective on the future of work. https://www.ilo.org/publications/major-publications/mind-ai-divide-shaping-global-perspective-future-work

Marcus, G. (2024, March 27). Don’t stifle AI with regulation. Newsweek.

https://www.newsweek.com/dont-stifle-ai-regulation-opinion-1980528

UNCTAD. (2025). Technology and Innovation Report 2025: Inclusive Artificial Intelligence for Development. United Nations Conference on Trade and Development.

https://unctad.org/system/files/official-document/tir2025_en.pdf