Funded projects

The AI4CI Hub’s Rapid Response Fund offers an exciting opportunity for small-scale, high-impact projects that unite academic and non-academic partners from across the UK. Discover the exciting projects currently under way and see how they’re turning new ideas about AI for collective intelligence into real-world impact.

Understanding bias for organisational ESG behavioural profiling

Project lead: Rudy Arthur, University of Exeter

Project partner: ekoIntelligence

The ekoIntelligence platform uses AI to analyze publicly available data—from news and corporate reports to academic research and NGO publications—identifying Environmental, Social, and Governance (ESG) trends. It creates clear maps of sustainability and risk for companies and investors. However, input data can be biased, amplifying certain narratives and omitting key facts, like who funded the research. This project uses advanced AI to assess the reputation of sources, detect these biases to provide more robust confi dence ratings. By scoring reliability and revealing influence networks, ekoIntelligence ensures users can trust the information they see, enabling better decisions about sustainability and risk.

Using collective intelligence to model the cryptocurrency circular-economy

Project lead: Padraig Corcoran, Cardiff University

Project partner: Office for National Statistics

A cryptocurrency is a decentralised digital currency that is not issued or controlled by a centralised authority. Initially, cryptocurrencies were primarily used by individuals as a store of value. However, more recently, many individuals have started to use cryptocurrencies as a medium of exchange to pay for goods and services. In some regions, adoption has progressed so much that there exists a corresponding circular-economy where cryptocurrencies are the primary medium of exchange. In this project, we will examine how combining AI with crowdsourcing can be used to model the growth of this circular economy.

GNOME (Graph Neural Ordinary Differential Equations for Modelling Epidemics)

Project lead: Daniela De Angelis, University of Cambridge 

Project partner: UK Health Security Agency

During the COVID-19 pandemic mathematical models played a crucial role, providing predictions of disease burden (e.g. mortality) to inform policy decisions. However, existing models rely on strong assumptions about how infection spreads. They also lack geographical resolution, resulting in predictions that only cover large areas and populations, of limited value in informing interventions at local level and for specific communities.

This project will develop new, flexible models, combining epidemiological understanding and machine learning to forecast disease burden at a fine geographic resolution. The increased flexibility and realism will improve predictions, leading to more reliable, targeted and actionable evidence for policy.

WARM-AIR: Wearable assisted respiratory monitoring for integrated recognition of COPD exacerbations

Project lead: James Dodd, University of Bristol

Project partner: my mhealth

In the UK, 1.7 million people have a lung disease called Chronic Obstructive Pulmonary Disease (COPD), which makes it hard to breathe and causes coughing. Sometimes, their symptoms suddenly get much worse, called “flare-ups.” These flare-ups make breathing even harder and send over 100,000 people to hospital every year. We want to use Artificial Intelligence (AI) to predict these flare-ups before they happen. By using data from 50,000 people using a COPD app and 900 people wearing smartwatches, AI could predict these flare-ups days in advance, giving people time to take action and avoid becoming seriously unwell.

CC4I: Creative Crowdsourcing For Innovation

Project lead: Sarah Jones, University of London

Project partner: Innovation Service Network

Computerised creativity models have the potential to support people to use information from multiple sources to recommend new ideas that improve collective decision-making in innovation processes. To explore this potential, the CC4I project is extending one such model that codifies established creative processes and outcomes using symbolic and non-symbolic AI techniques. The extended model manipulates information from multiple stakeholder groups to discover ideas across and in-between spaces of possibilities defined by information from these groups. It will be integrated into an established innovation crowdsourcing platform to be used in pilot projects to explore its eeect on collective and creative decision-making.

Embedding EDI in the distribution of research funding: an AI-assisted collective intelligence approach

Project lead: Evangelos Pournaras, University of Leeds

Project partner: Complex Chaos, Newlands Coaching and Consulting Ltd

This project will transform the distribution of research funding by redesigning a fairer, more participatory process for reviewing and selecting projects to fund, combining collective intelligence with Equality, Diversity and Inclusion (EDI) principles. Using rigorous methods of computational social choice and artificial intelligence, we will introduce novel preference elicitation and aggregation rules for research funding distribution to achieve a more proportional and diverse representation of both applicants and reviewers. We will test these innovative methods in real world in synergy with key academic and industrial partners working on systemic changes for EDI and scaling up collective human collaboration.

Participatory futuring of AI enabled town centres

Project lead: Alex Taylor, University of Edinburgh

Project partner: Architecture and Design Scotland, Scottish Borders Council

By prototyping a participatory approach to engaging with local communities, this six-month project will explore how AI4CI can be harnessed to enable Scotland’s Town Centres to become vibrant, healthy and creative places to live, learn and work in, and for all to enjoy and visit. Through three curated workshops focused on Town Centres in the Scottish Borders, we will engage local communities and use their collective insights and local knowledge to co-produce options for the future of the town. In this way, we will provide a transferable methodology for AI enabled Town Centres which accounts for geography and rurality.