Report: How Public Money is Shaping the Future Direction of AI: An Analysis of the EU’s Investment in AI Development

The EU has promised to invest €20bn in the “digital decade” until 2030, and announced the ambition to be the world’s “home of trustworthy Artificial Intelligence”. This research conducted by Eticas and commissioned by the European AI & Society Fund unpacks whether the available EU funding mechanisms are used strategically to support the political commitment to lead on human-centred, responsible and trustworthy AI.  The comprehensive research analyses how the EU funding is allocated, who receives it and how values and ethics that the EU aspires to follow are guiding its funds.  

The recommendations presented in this report include publicly accessible data, effective evaluation of the real-world impacts of funding, and mechanisms for civil society participation in funding before investing further public funds to achieve the EU’s goal of being the epicentre of trustworthy AI.

 Download the full report here.  

Key Findings 

• Despite a commitment to AI that works for people and is good for society, over 20% of calls for proposals involving AI from 2007-2020 were allocated to programmes focused purely on technological development without a clear area of application. In spite of a pledge to use technology to address the climate crisis, only 1.1% of the total calls related to the environment involved AI. 
• Although the framework programmes involve a rigorous ethics review and specific guidelines exist around ethical AI, only 30.3% of funding calls related to AI make any mention of issues of trustworthiness, privacy or ethics. Instead of integrating horizontally across calls, these issues are are siloed as separate areas of study. 

• Civil society organisations are not involved in the design of funding programmes, nor are they incentivised in any way to participate in the funding applications. It leads to question whether and how public interest is represented in AI development.  
• The EU does not conduct an impact assessment to evaluate the societal and economic impacts of the funded projects on AI. Despite considerable time being invested in monitoring and evaluation, the oversight mechanisms focus on the fulfilment of contracts, and not on the impact and actual results of the funded projects. 
• 73% of the available money from the FP7 and Horizon2020 programmes between 2007-2020 went to education institutions and research organisations. While private for-profit entities, including SMEs and start-ups were significantly underrepresented. 

Use our data for research 

Eticas has made the dataset publicly available for researchers and interested groups to interrogate and use the data in their own work. Lack of comprehensive and accessible data on EU’s funding flows and its implementation was one of the main hurdles Etica’s researchers faced when conducting this work. Therefore we decided to publish the dataset to address the opacity of EU funding for AI, and enable future research in this field.