AI Costs Lead Westpac to Prod Staff Toward ‘Sensible’ Model Use
TL;DR
Westpac Banking Corp. is tightening cost control around AI use and tracking token consumption across the company. Simpler tasks are being pushed toward cheaper models instead of defaulting to the most expensive frontier systems. The bank frames this as sensible model selection: premium AI for complex work, smaller models for routine prompting. The move fits a wider enterprise shift from broad AI rollout to workload-by-workload cost and value management.
Nauti's Take
Westpac looks less like an AI skeptic and more like an operator doing the math. The first AI phase rewarded experimentation, tool sprawl, and defaulting to the strongest models.
The next phase is more practical: which tasks really need frontier models, which are fine on cheaper systems, and who measures the difference? That discipline separates AI theater from AI operations.
Briefingshow
The story shows where many companies land after the first wave of AI enthusiasm: usage alone is no longer a success metric. Once every prompt has a real token cost, AI governance needs routing rules, budgets, and quality thresholds. For teams, the best model is not automatically the most expensive one, but the one that fits the task.
Sources