Daniele Bifolco

His research focuses on the modeling and management of intellectual property and privacy in AI-intensive software systems. He combines empirical software engineering and mining software repositories to study the provenance, licensing, and dependencies of software and AI components.

In particular, he investigates how code generated by large language models can be linked to existing source code and its associated licensing information. His work also examines the accuracy of dependency information provided by software hosting platforms and supports the generation of Software and AI Bills of Materials, which document software dependencies and AI components. These activities include contributions to CodeGenLink, a tool for identifying the likely origin and license of generated code, and ALOHA, an AI Bill of Materials generator for Hugging Face models.