The chief executive of Commonwealth Bank of Australia (CBA), Matt Comyn, has warned that businesses are facing an emerging challenge as artificial intelligence becomes more deeply integrated into corporate operations: rapidly rising costs and an increasing volume of low-quality AI-generated work. Speaking at a business conference in Sydney, Comyn said many organizations underestimated how expensive AI would become once it moved beyond simple pilot projects into large-scale production use.
A major concern is the way enterprise AI services are priced. Unlike most consumers who use AI through fixed-fee subscriptions, businesses often pay based on the number of tokens—or units of text and computation—processed by AI systems. Comyn noted that costs do not increase in a predictable linear manner. As companies use more advanced AI models with reasoning capabilities, tool integration, and larger contextual windows, token consumption can rise dramatically, causing operating expenses to grow far faster than originally expected.
Comyn also criticized what he called “work slop,” referring to low-value AI-generated content that enters corporate workflows without adequate quality control. This can include reports, presentations, emails, code, or analyses that appear polished but contain inaccuracies, weak reasoning, or limited business value. He argued that organizations risk creating large volumes of content that require additional review and correction, potentially offsetting some of the productivity gains AI is supposed to deliver.
The comments are notable because CBA is one of Australia’s most active corporate adopters of AI. The bank recently hosted an AI summit featuring Sam Altman and appointed what it described as Australia’s first chief AI scientist within the banking sector. Rather than rejecting AI, Comyn’s message was that companies need to become more disciplined about measuring return on investment, controlling costs, and ensuring that AI-generated output genuinely improves business performance. His remarks reflect a broader shift across the corporate world, where enthusiasm for AI is increasingly being balanced with concerns about economics, governance, and practical value.
