Open Future

We are part of NGI Commons, a European project that is developing Europe's long term strategy for digital and internet commons. This term covers a broad range of open assets and digital public goods, including open-source software, open hardware, open design, open licensing, open standards, and open data. The project is built on top of the NGI initiative, which is a globally relevant example of large-scale public investment in digital public goods and related open assets. Open Future’s role in the project is to develop a strategic agenda for Digital Commons policies, with a focus on financial frameworks. Our aim is also to increase awareness and promote various types of open source and digital public goods initiatives. As part of the project, we are looking more closely at commons-based initiatives, that is, those where digital public goods are created by initiatives that are collectively or communally governed.

Public AI is a term that covers policy proposals and frameworks that support the development of AI systems in the public interest. Open Future has written, with partners, one of several briefs that provided an initial framing on how public AI can help democratise AI development and access. In the next 12 months, we will be working on a new policy brief that defines pathways to developing public AI systems in more detail. This includes exploring ways to support the open provision of AI training datasets and AI models, which ideally fulfil the DPG Standard. We are also working with partners to develop a set of principles aimed at funders and policy-makers in the AI development space, which will ensure that AI systems are developed in the public interest and openly shared. The concept of public AI aligns closely with the vision of AI as a digital public good.

Over the last three years, we have been conducting work on data commons: approaches to data governance based on the principles of the commons, including open data. We have published a range of white papers and primers on data commons, including the latest one, "Commons-based data set governance for AI". In the coming 12 months, we aim to build on this work, with the goal of bringing to life a vision of datasets shared as digital public goods. To this end, we are working with the Open Source Initiative to develop recommendations for open source AI developers on how more data and datasets can be shared as digital public goods. We are also continuing to analyse challenges to data sharing related to concentrations of power, and the need to protect data commons - for example by analysing various open licensing approaches. Finally, we are aiming to apply the data commons approach to practical cases of open source AI development.