Decentralized AI

Decentralized AI is one AI Sweden’s strategic initiatives. While centralized AI systems with access to all data and information in the cloud or in a single device are easier to engineer and implement, decentralised systems are becoming increasingly important not least due to data privacy restrictions and limited bandwidth.

Decentralized AI will play a critical role in the use of AI in society. It is the technology behind everything from hospitals to autonomous cars being able to share and benefit from knowledge centrally, while keeping the sensitive data safe and local.

The initiative consists of developing an open lab environment, prioritizing high-impact use cases, and solutions to both core technical issues and legal challenges. We encourage robust thought leadership in the field and we boost joint research projects together with partners. This will accelerate Sweden’s application of AI across sectors in a secure and efficient way.

Join us in accelerating the development of decentralized AI. Our educational videos are a great way to get started.

Edge Learning Lab

Edge Learning Lab gives AI Sweden’s partners the chance to position themselves at the forefront of Edge AI and federated learning by scoping projects together and quickly build a working environment for experiments.

Edge Learning Lab

Projects

Federated Learning In Banking

Federated learning for anti-money laundering in banking

SpaceEdge

A testbed for application development in space

Regulatory Pilot Testbed

DataRätt InnoVation (DRIV)

Få koll på juridiken i ditt AI-projekt

What is decentralized AI?

Decentralized AI refers to moving intelligence and learning out to different devices and organizations. We can train machine learning models on locally available (i.e. decentralized) data and make local decisions. The approach enables combining knowledge from several local datasets, without distributing the actual raw data between devices, locations, and organizations. Unlike algorithms trained on a centralized dataset, decentralized learning distributes models rather than the data itself.

Decentralized and federated learning approaches are quickly becoming one of the most important areas of applied AI, mainly due to efficiency and privacy reasons. It is a prerequisite for being able to apply AI on a broad scale in Sweden. In real world scenarios, data is often impossible to centralize in one single data center. As a result, if we want to make the most of all data generated in society, we will have to adopt decentralized approaches to AI and learning. Federated learning and Edge AI are two sub-areas of decentralized AI that are being researched by AI Sweden partners.