SECure: A Social and Environmental Certificate for AI Systems
By Abhishek Gupta, Camylle Lanteigne, Sara Kingsley. 2020
16 citations recorded by the source corpus.
In a world increasingly dominated by AI applications, an understudied aspect is the carbon and social footprint of these power-hungry algorithms that require copious computation and a trove of data for training and prediction. While profitable in the short-term, these practices are unsustainable and socially extractive from both a data-use and energy-use perspective. This work proposes an ESG-inspired framework combining socio-technical measures to build eco-socially responsible AI systems. The framework has four pillars: compute-efficient machine learning, federated learning, data sovereignty, and a LEEDesque certificate. Compute-efficient machine learning is the use of compressed network architectures that show marginal decreases in accuracy. Federated learning augments the first pillar's impact through the use of techniques that distribute computational loads across idle capacity on devices. This is paired with the third pillar of data sovereignty to ensure the privacy of user data via techniques like use-based privacy and differential privacy. The final pillar ties all these factors together and certifies products and services in a standardized manner on their environmental and social impacts, allowing consumers to align their purchase with their values.
Subfields: Safety Research, Safety, Robustness & Interpretability
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Safety Research
In Safety, Robustness & Interpretability
- Anne-Laure Ligozat
- Ben Shneiderman
- Marina Jirotka
- Jonathan Brophy
- J. Konstan
- Alexandra Sasha Luccioni
- Daniel Lowd
- Battista Biggio
Citations in the corpus
- Measuring the Carbon Intensity of AI in Cloud Instances (2022, neutral): “Building upon this research, efforts have also been made to certify systems as being socially- and environmentally-conscious [14], working towards comparing both the environmental costs and potential benefits of AI models in order to paint a more holistic picture of AI.”
- Machine Unlearning for Random Forests (2020, background): “…Finally, this line of research promotes a more economically and environmentally sustainable approach to building learning systems; if a model can be continuously updated only as necessary and avoid frequent retraining, significant time and computational resources can be spared (Gupta et al., 2020).”
- Estimating the Carbon Footprint of BLOOM, a 176B Parameter Language Model (2022, background): “, 2021) and the certification of ML systems according to their social and environmental impacts (Gupta et al., 2020).”
- DART: Data Addition and Removal Trees (2020, unknown): “We believe our work not only has positive implications for user privacy, but also opens up new possibilities for continuous model learning while having a positive impact on the environment (Gupta, Lanteigne, and Kingsley 2020).”
- Six Human-Centered Artificial Intelligence Grand Challenges (2023, background): “, a lighter social and carbon footprint) (Gupta et al., 2020); otherwise, humans risk grave environmental and social harm and not having genuine social wealth (Ghosh, 2015) and a viable future.”