AI and climate change – 13 policy recommendations to meet current challenges
Artificial intelligence (AI) can be a powerful tool in the fight against global warming. The EU Commission sees the fields digitization and climate as closely intertwined.
This is why Commission President Ursula von der Leyen also refers to this strategic link as “Twin Transitions” that will have a significant impact on our future. Just recently, its importance has been underlined again by the conditions attached to the European Recovery Fund in the wake of the COVID-19 pandemic. These demand that a large part of the funds must be directed towards solving the future tasks of climate and digitization.
The long-term goals are clear, but what are the pitfalls along the way that must be avoided, specifically in the development and use of AI? The policy paper “The AI Gambit: Leveraging artificial intelligence to combat climate change: opportunities, challenges, and recommendations” analyses these questions and translates the findings into concrete policy recommendations.
The publication is the result of a collaboration between the Vodafone Institute and a research team of Oxford Internet Insitute consisting of Prof. Luciano Floridi, Prof. Mariarosaria Taddeo, Josh Cowls and Andreas Tsamados, the two main authors of a research paper on which this summary is based.
In addition to opportunities, the paper focuses primarily on two challenges that arise in the context of the climate issue. One concerns social and ethical challenges of AI, such as bias, discrimination, or non-transparent decision-making.
The other is more climate-specific: what is the energy demand of AI developments and what is the contribution of computationally and training intensive AI systems to global climate change through their CO2 emissions? The current lack of public data makes it difficult to measure these. The authors overcome this by looking at the carbon footprint of AI research only to present an example that provides important insights for the discussion.
The summary of the research paper offers an overview of the main research findings and provides 13 policy recommendations towards policy makers and developers on how to leverage the opportunities of AI while minimizing the impact on the environment from the beginning.
The complete research paper can be found here, it is currently under peer review.