By Karli Kalpala, Head of Strategy & AI Agent Business at Digital Workforce
For the first time in human history, we’re seeing technological change beyond new software or automation tools. For those of us who have spent the last decade inside large-scale automation and transformation programmes, the shift is unmistakable. Generative AI, reasoning models, and autonomous agents are emerging, reshaping how industries create value and generate and apply knowledge.
This is not simply another productivity wave. It represents a structural change in how work itself is organised.
What can we learn from 2025?
The conversation around AI in 2025 has shifted from hype to acceptance. The technology is no longer experimental, and access to AI tools is no longer a differentiator. Most organisations now have some form of generative AI capability, and large institutions, including major banks, have already introduced generative-AI-based agents and solutions to their end customers. We have clearly passed the tipping point.
What distinguishes organisations now is not access to AI, but the ability to operationalise it safely, repeatedly, and at scale. Looking ahead to 2026, businesses must increasingly prioritise workforce readiness: ensuring employees are equipped to work effectively alongside AI tools and autonomous agents as part of everyday operations.
Research shows that only 8-10% of businesses are truly AI-ready. As competitive pressure increases, the gap between those that can translate AI into operating advantage and those that cannot will widen rapidly.
Why AI pilots have stalled, and what to do differently
The main barrier keeping companies from scaling successful AI pilots into full deployment is a misunderstanding of what real transformation requires. Take Gartner’s prediction that over 40% of agentic AI projects will be cancelled by 2027, and MIT’s discovery of a 95% failure rate at enterprises for generative AI pilots.
But when examined closely, these figures do not indicate a failure of technology. The tools work. What fails is the surrounding system: the workflows, incentives, skills, and operating models that determine whether AI can be used effectively in production.
In practice, around two-thirds of the effort required to deliver bottom-line impact sits in familiar territory: redesigning workflows, redefining roles, managing risk and controls, and helping people adapt to new ways of working. Pilots and proofs of concept built in isolation rarely deliver lasting value because they are not designed to change how work actually happens.
Successful transformation requires integrating AI into live operations, not experimenting with it on the sidelines.
What’s driving AI acceleration?
One of the most important drivers of acceleration over the coming year will be the operationalisation of reasoning models. Until recently, information has primarily flowed between people, or between people and documents. Reasoning models change this dynamic by enabling systems to interact directly with documents, data, and tools without requiring constant human intervention.
In regulated environments such as financial services, we are already seeing early examples of this shift. Autonomous agents can coordinate document analysis, data validation, and decision support across systems, provided governance, controls, and human oversight are designed in from the outset.
The implication is significant: we can now begin to decouple outcomes from human labour in knowledge-intensive services that were previously out of reach of automation, including areas such as asset management, accounting, insurance, and legal services.
What differentiates the organisations that are leading in 2026?
The idea that most generative AI projects fail misses the real point. AI tools open up possibilities that did not exist before, but technology alone does not create value. Real impact comes from how organisations redesign work and prepare people to operate differently. This is a revolution of Work itself and who / what performs it in our organisations not that much about the software architecture.
Sandbox projects rarely produce measurable results because they are not built to reshape live operations. Inflated expectations came from assuming that deploying AI would automatically deliver benefits, when success still depends on leadership, commitment, and a willingness to change how decisions are made and work is performed.
Organisations that lead in 2026 will ensure their teams are AI-literate across departments and workflows. This goes beyond basic tool usage. It means enabling people to collaborate effectively with autonomous agents, redesigning exception handling and oversight models, and building confidence in how AI systems are governed and monitored.
This is a defining moment for the financial services sector, with large institutions and organisations in the sector having been the most visible AI movers this year, particularly in deploying customer-facing AI solutions.
Is now the time for me to invest in AI initiatives?
There has been speculation about an AI bubble bursting in 2026, but this view overlooks a familiar trajectory. With any transformative technology, we see an initial surge of investment followed by a period of correction. What is often described as a “bubble bursting” is more accurately a shift toward maturity. It is no wonder why Gartner’s hypecycle has the shape of an S-curve.
In 2026, organisations will become more disciplined about proving value before scaling. This is a healthy evolution. It will lead to fewer superficial initiatives and more deployable, operationally grounded AI solutions that can be scaled with confidence.
We are better equipped than in previous technology cycles. The lesson for financial services and other regulated industries is clear: speed of adoption and return on investment are not determined by technology alone.
Organisations that want to move beyond simply ticking the AI box must recognise that the majority of the work lies in people and processes. Failing to prepare the workforce for the most significant technological shift this industry has seen will leave even the most advanced AI investments under-delivering in 2026. In the next phase of AI adoption, technology will be assumed — readiness will be decisive.
About the Author
Karli stands among the most senior experts in strategic business process automation within the Nordics, having worked in the industry since its inception. Today, Karli holds the Head of Strategic Transformation position at Digital Workforce. His work includes supporting some of Europe’s largest banks and financial services organizations in optimizing strategic business operations and building comprehensive solutions spanning the customer journey.
