The democratization of artificial intelligence (AI) has been a central theme in AI research over the past decade, yet no systematic analysis has examined how this concept has evolved alongside emerging technologies such as generative AI and low-code or no-code platforms. This study examines the decadal evolution of AI democratization research from 2017 to 2026. A total of 427 unique papers were analyzed from Scopus and Web of Science databases. Keyword frequency, annual growth rates, and h-index metrics were examined across three distinct eras: Foundation and Accessibility (2017-2020), Technical Maturation and Participation (2021-2024), and Generative Democratization (2025-2026). Ranking of 20 core keywords was tracked over the ten years. Artificial intelligence remained the dominant keyword with 108 documents and an h-index of 16. Generative terms showed the highest annual growth rates (AGR) (large language models at 0.5 AGR) but low citation impact (h-index of 2). Democratization-related keywords maintained a steady presence with moderate h-indices of 3 to 4. AI literacy climbed from rank 13 to rank 9 by 2025-2026, while no-code and low-code tools collectively matched the volume of deep learning research.The findings reveal that the generative surge has not replaced AI democratization but rather been layered with new priorities, including governance, participation, and literacy. The transition from building models to governing them suggests that future research must focus on user education and ethical stewardship. True democratization now requires not just access, but the ability to understand, control, and meaningfully participate in an AI system.

JEL Codes:

Keywords: AI democratization, artificial intelligence, no-code AI, low-code AI, AI literacy