74% of AI Value Goes to 20% of Firms. They’re Not Cutting Costs.

Most Australian businesses that use AI point it at the same problem: cut costs, save time, automate the repetitive stuff. PwC’s 2026 AI Performance Study, released on 13 April and covering 1,217 executives across 25 sectors, says that approach leaves most of the available value on the table. The companies capturing real returns are pointing AI at growth. The gap between the two groups is not small.

The headline: 74% of all AI-driven economic value is being captured by just 20% of organisations. The top fifth generate 7.2 times more AI-driven financial performance — measured as revenue gains and efficiency improvements adjusted against industry medians — than everyone else. That’s not a rounding difference. That’s a chasm.

The differentiator isn’t technology spend or the number of AI tools deployed. It’s intent. AI leaders are 2.6 times more likely to use AI to reinvent their business model and two to three times more likely to pursue revenue opportunities across sector boundaries. They’re also twice as likely to track AI’s business impact systematically. The firms stuck in the bottom 80%? They’re running AI on efficiency — drafting emails faster, automating data entry, trimming admin hours — without connecting those gains to growth.

PwC’s Global Chief AI Officer Joe Atkinson was direct: “Many companies are busy rolling out AI pilots, but only a minority are converting that activity into measurable financial returns. The leaders stand out because they point AI at growth, not just cost reduction.”

The PwC data lands on a pattern we’ve been tracking in Australian SMEs for months. Deloitte found only 5% of Australian SMEs extract real value from AI. Anthropic’s Economic Index showed Australian users over-index on admin tasks — management and office support — while under-indexing on the kind of strategic work PwC’s leaders prioritise. And Thomson Reuters reported that only 18% of professional services firms even measure AI ROI.

The picture is consistent. Australian businesses are adopting AI at world-beating rates — seventh-highest per capita globally, per Anthropic — but they’re using it to shave minutes off admin tasks. PwC’s study quantifies what that focus costs: seven-eighths of the available value, left uncaptured.

For a trades business, the cost-cutting version of AI is a chatbot that drafts quote follow-ups or scheduling software that trims drive time. Useful, but incremental. The growth version is a system that analyses job history to identify which customers are due for preventive maintenance, predicts seasonal demand spikes so you can staff ahead of competitors, or prices emergency callouts dynamically based on real-time capacity. Same underlying technology. Different aim. Different returns.

For an accounting firm, the cost-cutting version is document processing that closes the books faster. The growth version is AI that identifies which clients are outgrowing their current service tier, flags advisory opportunities from transaction patterns, or spots industries where your firm’s expertise could open an entirely new client segment. That’s revenue capture, not just admin leverage.

PwC’s data identifies three practices that separate the top 20%. First, they fund AI based on growth potential, not cost savings — and they invest 2.5 times more of their revenue into AI than peers, with monthly reviews to scale what works and kill what doesn’t. Second, they build trust infrastructure: leaders are 1.7 times more likely to have a documented responsible AI framework, and their employees are twice as likely to trust AI outputs. Trust isn’t a nice-to-have. It’s the mechanism that gets AI from pilot into operations. Third, they collaborate beyond their sector. Cross-sector collaboration is the single strongest factor in PwC’s AI fitness index — stronger than technology spend, headcount, or the number of tools deployed.

Pick the AI initiative in your business you’re most invested in. Ask whether it’s pointed at cost or growth. If the answer is cost — faster quotes, less admin time, cheaper document processing — ask the next question: what would this tool need to do to generate new revenue instead? It might be the same tool, deployed with a different aim. That’s the 7.2× gap in one decision.

Key takeaways

Is your AI pointed at growth or just at costs?

Most businesses start with efficiency. The ones pulling ahead have already redirected AI toward revenue. A short conversation is enough to identify where the growth lever sits in your operation. Book a call to talk it through.