# Generative Artificial Intelligence for Internet of Things Computing: A Systematic Survey

**Authors:** Fabrizio Mangione, Claudio Savaglio, Giancarlo Fortino  
**arXiv:** [2504.07635](https://arxiv.org/abs/2504.07635)  
**Date:** 2025-04-10  
**Summary page:** https://www.summarizepaper.com/en/paper/2504.07635v1/

## Key points

- Integration of Generative Artificial Intelligence (GenAI) within the Internet of Things (IoT) has garnered significant interest in sectors such as healthcare, manufacturing, and smart cities.
- Extensive research is being conducted to explore the potential synergy between GenAI and IoT to innovate and reshape existing paradigms.
- A tripartite framework was employed in a systematic survey to categorize and analyze literature results into network-, object-, or semantic-oriented domains.
- Categorization highlights distinct features, overlapping aspects, transformative potential of IoT, challenges, and research opportunities within each domain.
- Critical trends, gaps, and emergent opportunities were identified across dimensions to inform theoretical exploration and practical implementation strategies.
- Recent studies adopting GenAI methods in internet-, object-, and semantic-oriented IoT scenarios were mapped to key research questions focusing on application areas, applied DGMs, key challenges faced, and proposed future directions.
- The study aims to address gaps in existing research by providing valuable insights for experts and newcomers alike.

## Summary

The integration of Generative Artificial Intelligence (GenAI) within the Internet of Things (IoT) has garnered significant interest due to their continuous evolution and widespread adoption in various sectors such as healthcare, manufacturing, and smart cities. This convergence has sparked extensive research to explore the potential synergy between GenAI and IoT, aiming to innovate and reshape existing paradigms. In this systematic survey, a tripartite framework was employed to categorize and analyze a high number of literature results. Articles were categorized into network-, object-, or semantic-oriented domains to highlight distinct features and overlapping aspects. This categorization underscores the transformative potential of IoT while identifying challenges and research opportunities within each domain. By synthesizing insights across these dimensions, critical trends, gaps, and emergent opportunities were identified to inform theoretical exploration and practical implementation strategies. The analysis provides a comprehensive view of IoT's trajectory by mapping recent studies adopting GenAI methods in internet-, object-, and semantic-oriented IoT scenarios. Each article was mapped to key research questions focusing on application areas, applied DGMs according to taxonomy introduced earlier, key challenges faced, and proposed future directions. The study aims to address gaps in existing research by providing valuable insights for experts and newcomers alike. Overall, this systematic survey offers a holistic overview of the opportunities, issues, and considerations arising from the integration of GenAI within the broader IoT ecosystem. It highlights critical trends shaping the future landscape of GenAI-IoT integration while paving the way for further advancements in this evolving field.