Generative artificial intelligence has moved from novelty to near-ubiquity in commercial settings, according to recent research — but widespread use is not the same as informed, governed or value-driven deployment. That distinction is now the central challenge for business leaders who must decide where AI can genuinely cut costs, boost productivity or change services without unintended consequences for employees, customers or the balance sheet.
Scale of adoption
A 2026 survey compiled in the Stanford AI Index found that 88% of organisations reported using AI in at least one business function in 2025, while 70% said generative AI was in use somewhere in the business. Those headline figures underline how quickly tools such as chatbots and copilots have become embedded in everyday operations, but they do not reveal whether applications are strategic or experimental.
| Measure | Share of organisations (2025) |
|---|---|
| Any AI in at least one function | 88% |
| Generative AI in at least one function | 70% |
From tool use to organisational understanding
Firms now confront a transition from token use — asking an assistant for draft text or automated summaries — to embedding models into processes that affect hiring, pricing, customer relationships and compliance. Successfully making that leap requires more than familiarity with interfaces: it demands leaders who grasp where models add value, what data they should see and how to maintain human oversight.
Executives are being pushed to treat these questions as strategic, not simply technological. Decisions around which problems to automate, how to evaluate model outputs and what safeguards to place around sensitive information are also business judgements about risk appetite, cost control and reputation.
What leaders should check now
Organisations should carry out a rapid readiness assessment to align capability with governance and measurable outcomes. Practical questions include whether there are defined business outcomes tied to AI projects, whether data pipelines and governance are mature enough for model training or inference, and whether staff possess sufficient AI literacy to scrutinise outputs and risks.
- Do AI projects have clear, measurable business objectives?
- Are data flows and governance structured for safe model use?
- Can teams assess outputs and spot errors or bias?
- Are there policies for selecting and monitoring models?
- Is human oversight specified for high-risk decisions?
Those are not merely technical checkboxes. They reflect workplace choices that will affect jobs, pay and the allocation of tasks between humans and machines. Poorly governed AI could create operational blind spots that damage customer trust or trigger regulatory scrutiny, while thoughtful deployment can cut routine costs and free staff for higher-value activities.
Implications for wages, prices and jobs
When AI displaces lower-value tasks, businesses may lower operating costs and press margins — a dynamic that can put downward pressure on prices or allow reinvestment in growth. Equally, firms that invest in upskilling managers and analysts to use AI responsibly may capture productivity gains without immediate headcount reductions. The distributional effects will depend on whether employers pair automation with retraining and clear governance.
Policymakers and company boards should therefore insist on transparency around outcomes and workforce plans. Without that, the rush to exploit generative tools risks concentrating benefits with early adopters while leaving many workers exposed to rapidly changing role definitions and oversight gaps.
The rapid adoption rates reported by the Stanford index mark a turning point for corporate strategy. The next, harder step is turning that adoption into durable, governed value — a task that will determine how AI changes pay packets, prices paid by consumers and the structure of many jobs.
