This year will be a decisive turning point in how organisations deploy and derive value from artificial intelligence (AI). This is according to Cliff de Wit, Accelera Digital Group (ADG) Chief Innovation Officer.
De Wit says businesses will move beyond experimentation this year. They will graduate into true operationalisation, with AI becoming a dependable and embedded driver of productivity, and not just an overhead.
“For the past few years, many organisations have been stuck in the AI tax phase, spending more time fixing AI than using it and struggling to measure its return. In 2026, that changes as AI becomes operational, measurable and genuinely transformative,” he says.
This shift is reinforced by global data, with Google Cloud’s ROI of Gen AI report showing that 74% of organisations already see return on investment (ROI) from their generative AI investments. A majority (86%) of companies reporting revenue growth attribute increases of 6% or more directly to AI.
De Wit highlights four critical shifts business leaders must embrace to unlock this next phase in their own companies.
From ‘chat’ to agentic workflows
First, the era of trying to make chatbots useful is ending.
“We’re shifting from conversing with bots to deploying agents trained to perform tasks effectively,” he comments.
Organisations will increasingly use autonomous, task-oriented AI agents to execute complex workflows. As De Wit explains, instead of asking a bot to help review invoices, you will instruct an agent to: “Review these 500 invoices and alert me to any discrepancies against our VAT policy.”
This marks a shift from prompting to orchestrating, and will reshape teamwork.
As agentic workflows free employees from repetitive tasks, they can focus on higher-value decision-making and enhancing organisational productivity.
Prioritising AI-first processes to remove legacy tax
Many organisations have attempted to retrofit AI into outdated processes, often compounding inefficiencies. De Wit warns that this approach is no longer viable.
“Legacy processes accrue more tax every time you try to bolt AI onto them. You end up with even more layers on top of already bloated workflows. In 2026, leaders must look for opportunities to design AI-first processes from the ground up,” he states.
That means re-examining entire value chains and identifying where AI can eliminate steps entirely. This is because in some cases agentic AI approaches not only optimise a process, but also remove the need for certain steps altogether.
Measuring what matters: Real ROI, not adoption rates
The metrics that matter are shifting as AI deployments mature. De Wit emphasises that adoption numbers, once considered the primary measure of success, are no longer sufficient.
“Organisations must focus on real metrics, like quantifiable productivity improvements and cost reductions. If an AI initiative does not show value within a few months, cut it,” he says.
The ROI of Gen AI report supports this stance, showing that:
· 63% of organisations report business growth directly driven by AI;
· 45% of organisations with productivity gains say employee productivity has at least doubled; and
· 91% of organisations with strong C-suite sponsorship report revenue increases of 6% or more.
The talent shift: From prompting to orchestrating
The rapid evolution of AI capabilities is also reshaping talent requirements. Prompt engineering dominated early AI adoption, but De Wit believes that a new skillset will define the next phase.
“In 2026, the ability to orchestrate agents will be even more important than writing the best prompts. Users are learning when and how to use AI and, crucially, when not to trust it blindly,” he says.
This shift requires a blend of technical understanding, process design and critical thinking.
“We are moving into a world where employees must be comfortable managing AI-driven workflows, validating outputs and ensuring alignment with organisational policies,” De Wit says.
“Leaders who embrace agentic workflows, AI-first processes, real ROI metrics, and foster the new talent landscape will unlock extraordinary value.”
