Does Generative AI Create a “Creativity Tax”? Georgia’s Experience in a Global Context

In the Winter 2025 issue of MIT Sloan Management Review, the article “Does GenAI Impose a Creativity Tax?” by Francisco Castro, Jian Gao, and Sébastien Martin warns that the rapid spread of generative artificial intelligence may reshape not only productivity but the very nature of human creativity. Their main argument is that while GenAI undeniably boosts efficiency, it may also impose a “creativity tax”—a point at which individuals gradually abandon their personal style and rely instead on automatically generated “good-enough” outputs.

The article explains how this shift can reduce the diversity of ideas. When thousands of people use the same models, identical arguments, phrases, and structures begin to loop back into the training data, leading to a subtle homogenization of thought. The authors describe this as a potential homogenization spiral, where models learn from their own outputs and gradually lose originality. They argue that this can be mitigated only through deliberate organizational culture—where companies set rules ensuring human contribution and require employees to interpret and reshape AI outputs rather than accept them uncritically.

In Georgia, the use of generative AI has grown rapidly during 2024–2025. Financial institutions, public agencies, and universities are actively testing AI models, but the local ecosystem is still searching for balance between speed and originality. The National Bank of Georgia’s AI Sandbox, launched in June 2025, became the region’s first regulatory laboratory allowing fintech firms to test algorithms in a controlled environment. While this initiative strengthens innovation and financial safety, it also raises a key question: can regulated experimentation preserve space for creative autonomy and design freedom?

Perhaps the most sensitive issue for Georgia is language resources. Currently, Georgian-language data represent only a fraction of global AI training materials, increasing the risk that Georgian content could be overshadowed by English-language narratives. To address this, the Enagram project launched to build open Georgian language datasets and corpora, ensuring that future AI models reflect the country’s cultural and educational realities.

BTU’s research (BTUAI 2025) indicates that many students using GenAI for coursework tend to submit unedited outputs, showing how easily efficiency can override originality. This points to the education system as the first line of defense: students should learn to treat AI-generated text as a starting point, not a finished product.

Neighboring countries are taking their own paths. In 2025, Armenia announced a $500 million investment in an AI Factory, aiming to build a regional innovation hub, while Azerbaijan adopted its Artificial Intelligence Strategy 2025–2028, emphasizing automation of government services and AI skills in education. Georgia, meanwhile, signed the Council of Europe Framework Convention on Artificial Intelligence, making it the first binding legal commitment to align AI with human rights, democracy, and transparency.

The creativity tax is not just a metaphor but a real risk for countries integrating GenAI into their economies and classrooms. As Georgia embraces this technology, it must preserve its linguistic identity, intellectual diversity, and cultural originality—so that it becomes not merely a user of AI, but a co-author in its development.

This article is based on MIT Sloan Management Review’s piece “Does GenAI Impose a Creativity Tax?” and incorporates insights from BTUAI’s 2025 regional research on AI ethics and creativity in the South Caucasus.