SECAI Core · phase 6 of 15
Sustainability
An AI principle that examines how efficiently resources are used and the broader effects of those usages
The Explain card
- Plain English
- Sustainability is the principle of examining how efficiently an AI system uses resources such as energy, compute, water and money, and the broader effects of that usage.
- Example
- A team chooses a very large model for a simple classification task. A smaller model would do the job at a fraction of the compute, cost and carbon, and respond faster too.
- Why it matters
- Resource waste has a security angle: oversized models and runaway inference loops are exactly what denial-of-wallet attacks exploit, and efficient systems are easier to monitor and afford to protect. Right-sizing is good engineering and good defence.
- Hook
- Sustainability: use the smallest model that does the job well.
Where it sits in the deck
Phase 6: Model Quality, Ethics, and Responsible AI Principles
Before deploying a model, understand the quality and ethical dimensions — bias, fairness, explainability, and the human oversight needed to govern outputs responsibly.
