Surviving the AI era is a business problem, not a technology one.

There is a comfortable story doing the rounds in Australian boardrooms. Adopt some artificial intelligence tools, train a few people, and the business will keep pace with the AI era. The data says that story is wrong. Adoption is not the same as survival, and most of the businesses adopting AI right now are getting very little back for it.

The evidence is stark. Research from MIT's Project NANDA in 2025 found that about 95 per cent of organisations are seeing no measurable return to their profit and loss from generative AI pilots. Only around five per cent capture value at scale. Separately, S&P Global Market Intelligence reported that 42 per cent of companies abandoned most of their AI initiatives in 2025, up sharply from 17 per cent a year earlier. Buying the tools is not the hard part. Turning them into results is.

What surviving the AI era actually means

Surviving the AI era is not a technology problem. It is a business problem wearing a technology costume. The RAND Corporation found that more than 80 per cent of AI projects fail to deliver their intended business value, roughly twice the failure rate of comparable technology projects. When researchers look at why, the causes are almost never about the model. They are about the business: a purpose no one agreed on, data foundations that were never built, executive sponsorship that faded after launch, and a habit of chasing the technology instead of the outcome.

Put plainly, the businesses that survive are not the ones with the best tools. They are the ones that treat AI as a transformation of how they operate, governed and measured like any other major commercial commitment.

The Australian picture: busy, not yet paid

Australian businesses are adopting quickly. The National AI Centre reported that about 43 per cent of Australian small and medium enterprises had some level of AI adoption across the December 2025 to February 2026 quarter, rising to 44 per cent in February, its strongest result in months. Adoption is broadening from one-off experiments into multiple business areas.

The return is a different story. Australian surveys show a large share of businesses reporting a positive impression of AI while close to half do not measure its impact at all, and only a small minority have built AI into the products and services they actually sell. That is a cost-saving reflex, not a growth strategy. It also leaves a business exposed, because a competitor who does treat AI as a growth strategy will out-run it.

There are also real barriers holding value back. The National AI Centre found that among businesses not adopting, around 65 per cent cited distrust of AI decision-making or a strong preference for human control, and 54 per cent felt AI was simply not relevant to them. Adoption without trust, governance and relevance is fragile. It stalls the moment something goes wrong.

What the survivors do differently

The roughly one in five organisations that get real value from AI are remarkably consistent. Five behaviours separate them from the rest.

• They define success before they start. A quantified target and a way to measure it are agreed before a project is approved, not reverse engineered afterwards.
• They build the data foundation first. Quality, access and governance are treated as the groundwork, not a problem to discover mid-project.
• They keep executive ownership alive. Sponsorship does not evaporate at launch. Someone senior remains accountable for the outcome.
• They treat AI as business transformation, not an IT project. The operating model, the roles and the process change with the technology.
• They govern it. Human oversight, output checking and clear accountability are built in, which is exactly what earns the trust that adoption needs to stick.

How DivineLab Worx helps businesses survive and grow

Our business consulting for the AI era is built on the same evidence-gated logic we apply to every engagement. We do not start with a tool. We start with the commercial question: where does AI actually create value in this business, what would prove it, and what has to be true in the data, the operating model and the governance for that value to be real. The output is a clear, board-grade view of where to act, where to wait, and how to measure the result before capital is committed.

This sits inside our AI advisory and governance and strategy and transformation work, and it connects naturally to capital-efficient growth, because for most established businesses the fastest AI return comes from applying it to the base you already have, not from a moonshot. If your business is adopting AI but cannot yet point to the return, that gap is the thing worth closing first.