Project’s overview
Using AI systems to mitigate and adapt to climate change can bring benefits, but it entails also negative consequences for the environment and the fundamental rights of people living near the infrastructures required to train and run AI models. The relationship between AI systems and the protection of the environment is twofold.
On the one hand, AI systems may be an enabler of sustainable development or climate change mitigation (so-called “AI for sustainability”); on the other hand, the use of AI models causes serious negative impacts on the environment (so-called “sustainability of AI”): i.e. alarming energy and water consumption, raw materials extraction’s negative consequences for non-EU countries’ territories, and e-waste disposal’s and recycling environmental concerns.
In the majority of cases, negative environmental impacts are externalised to non-EU countries, mostly in the Global South, even if the output generated by AI system is used in the Union. Similarly, the fundamental rights of people living near the infrastructures required to train and run AI models risk being violated.
However, a robust regulatory discourse and a roadmap that promotes an environmentally sustainable and eco-centric approach to AI is missing at the EU level.
EcoAI aims thus at creating a new EU legal framework for an ecological AI to align AI development with sustainability from an eco-centric perspective.
Methodology
EcoAI employs and combines different methodological approaches. Apart from desk research, the project entails conducting an online questionnaire (qualitative), semi-structured interviews and a focus group. The main objectives of the online questionnaire and the interviews are the following:
1) Investigate whether public and private stakeholders in the two different areas of research, i.e. EU and Brazil, use sustainable AI technologies.
2) In case of actors using sustainable AI technologies, understand how they apply the legal and ethical principles, and whether they have developed guidelines or private recommendations.
3) Understand why certain legal requirements have been implemented in practice, while others not.
4) Understand whether there are biases and/or discriminatory practices towards minorities and vulnerable groups.
Project’s objectives
The EcoAI’s objectives are:
1) developing a new European legal framework on an Ecological AI;
2) drafting a proposal for an arrangement on AI for the Public Good, aimed at strengthening cooperation between the EU and Brazil;
3) mapping the actors concerned about the use of AI systems to assess their need and knowledge about the consequences of the use of AI on the environment;
4) promoting awareness strategies among the interested stakeholders in the EU and Brazil to provide them with the appropriate means to use AI in a sustainable way.
EcoAI will develop an empirically driven theoretical model and propose concrete solutions to regulate AI in an ecological way. The proposed new EU regulatory framework for an ecological AI may also be used as a blueprint for an EU framework on new technologies more broadly and for a new legal framework on ecological AI in Non-EU countries.