Chapter 09 · Sustainability & AISustainability & AI

Narrow AI

Meaning statusEstablishedSource recordDirect document linkedWhy these are different

Definition

AI designed and trained to perform a specific task or a limited set of tasks, in contrast to hypothetical general intelligence.

References

Pearson / BerkeleyRussell & Norvig — Artificial Intelligence: A Modern Approach

This source provides part of the technical or institutional basis for the definition.

OECD.AIExplanatory memorandum on the updated OECD definition of an AI system (2024)

This source supports the explanation of how the term is applied, measured or governed in practice.

Overview

What it means

All deployed AI — from a deforestation classifier to a large language model — is narrow in the technical sense: capable within its domain, unable to transfer that competence broadly without adaptation. The term keeps expectations honest about what today's systems are.

How it is used

In sustainability applications, narrow systems do the practical work: forecasting solar output, reading metres, flagging methane plumes. Even general-purpose models are applied through narrow deployments with defined inputs, outputs and failure modes.

Why it matters

The narrow-versus-general distinction protects decision-makers from over-trusting systems outside their validated scope — a discipline that matters when AI outputs feed environmental compliance, safety or investment decisions.