Wisenut starts third year of government-backed federated AI project
The company will develop on-device model lightweighting and encrypted federated analysis under a Ministry of Science and ICT and IITP programme.
Wisenut said on November 3, 2026, that it has begun the third year of research on its decentralized federated AI (deFAI) development and demonstration project. The work is supported by South Korea's Ministry of Science and ICT and IITP as part of the SW Computing Industrial Original Technology Development Program.
The project's stated goals include model lightweighting technology that would allow small-scale computing on smartphones and wearables, and high-speed communication between central servers and local devices.
Wisenut also plans work on privacy-preserving federated analysis, which processes encrypted data without collecting it centrally, and on a global hierarchical learning model that compares and links central large models with small on-device models.
The company's CEO said federated computing will be a key factor in spreading AI agents into security-sensitive fields, according to a report by digitaltoday.co.kr.
Quick answers
What is Wisenut's deFAI project?
It is a government-supported decentralized federated AI development and demonstration project now in its third year of research.
Who supports the project?
South Korea's Ministry of Science and ICT and IITP support it through the SW Computing Industrial Original Technology Development Program.