← All insights ←

AI in corporate strategy: back to the greenfield

By · Published: 16 September 2026

StrategyDigital TransformationAI AdoptionBoard of Directors
AI in corporate strategy: back to the greenfield

First published on 16 September 2026 in the Network Briefs series of the SwissBoardForum.

Before the discussion about artificial intelligence turns to technology and governance, another question is worth asking: how would we build the organisation if we could start again today?

When I speak with boards and executive teams about artificial intelligence, one question comes up quickly: how should we approach this topic? It involves three levels: the strategy, its implementation, and the risks together with the responsibility of the board. This article deals with the first level. Those who start with technology and governance often go round in circles at the beginning and make no progress.

Resistance and uncertainty are nothing new

I have followed developments in information technology for more than three decades. At Sandoz AG I was among the first in Switzerland to sit the final apprenticeship examination as a technical draughtsman using CAD. At the same company I helped build the first TCP/IP networks and put the first email and web servers into operation. Since then I have shaped each phase of digital transformation and been accountable for it, up to artificial intelligence today.

In each of these phases the initial investment was high, whether in hardware, software, training or a series of failed projects. For a long time it remained uncertain whether the projects would work technically and deliver a return on investment. Finding the right moment was difficult. Those who started too early paid for the learning curve. Those who started too late paid for the investment backlog. On top of that came resistance, because the old technology and the established processes still worked.

One example: in 1992 the first CAD workstations for 3D modelling in plant engineering cost around CHF 300,000 each. Resistance among the design engineers was enormous. I still remember a production facility for an active pharmaceutical ingredient. Thanks to 3D modelling it was completed eight weeks ahead of schedule, and the investment broke even within a few weeks. But replacing the drawing board with a CAD system was not enough. Processes had to be adapted or rethought.

What sets this wave apart

The situation with AI today is similar. Anyone who layers AI on top of existing processes finds that those processes were never designed for this degree of automation. Roles, workflows, incentives and tools do not fit. Small efficiency gains are possible, but the economic effect remains modest.

What differs is the pace. According to a UBS estimate, ChatGPT reached around 100 million monthly active users in January 2023, two months after launch. TikTok needed about nine months to reach that figure, Instagram around two and a half years. In February 2026 OpenAI reported 900 million weekly active users. At this pace, a three-year strategy cycle comes too late.

Unlike past waves of transformation, this one reaches a broad range of organisations quickly wherever the following conditions are met:

  • The service consists mainly of processing information.
  • The client will be able to produce the result or carry out the task without help.
  • The service is billed by the hour.
  • The activity is not protected by an official licence, a liability regime or a signature requirement.

Where all four conditions apply, the pressure on the sector is high. This brings an uncomfortable dynamic. Once clients notice that the work has become cheaper, price pressure rises. Revenue can then only be maintained by bundling the service differently or by entering new business areas.

I therefore recommend starting with a fundamental question: how would we build the organisation if we could start from scratch today, on a greenfield site, with the new technological possibilities at hand?

A thought experiment: the fiduciary firm

A fiduciary firm keeps the books, prepares annual financial statements, handles payroll, VAT and tax returns, and takes on tasks relating to company formation, domiciliation and succession. The four conditions apply to differing degrees, but they apply. Many tasks can be standardised and therefore automated.

This raises the essential question: what is the core task of a fiduciary firm? Recording accounting data by the hour, or taking responsibility for financial statements, tax returns, certificates and the assessment of borderline cases? As a client, what matters to me is a dependable partner who takes responsibility up to a defined point. Data entry is not the main concern.

If clients take on these activities themselves, this looks at first like lost business. It need not be, because new opportunities arise. There will be fewer data entry errors, but more systematic errors that look plausible: in the allocation of accounts, in the treatment of VAT or in accruals. A layperson is unlikely to notice such errors. The control function therefore becomes more demanding.

How the environment changes

In this case, the framework conditions at municipal, cantonal and federal level are likely to change. The systems of businesses and authorities will be connected in future. Elsewhere this is already routine. In the United Kingdom, HMRC requires digital record keeping and submission via interfaces under ‘Making Tax Digital’. For VAT this has applied to all VAT-registered businesses since 1 April 2022. Returns are no longer filed through a portal but through compatible software, directly from the ERP or accounting system. Income tax has followed in stages since 6 April 2026. I have put such an interface into operation myself. The effort to build and maintain these interfaces should not be underestimated and requires specialist expertise. For individual clients this is expensive, so offering them such platforms is an opportunity. A new business area emerges: that of the platform operator.

In auditing, too, the degree of automation is likely to rise considerably. Instead of relying on samples alone, auditors can use AI to analyse and assess ever larger data sets, up to complete transaction populations.

An approach that holds across sectors

The same exercise can be applied to other sectors. Consider wealth managers whose clients will assemble their own portfolios; IT service providers whose clients suddenly develop software themselves; law firms whose clients can already answer many legal questions on their own.

Many of these business models are likely to change fundamentally. This forces every organisation to analyse its value chain and to focus on the client of the future. Expertise and experience will be in even greater demand after the AI transformation than they are today, but in different roles.

From vision to implementation

The greenfield thought experiment produces a vision of what the organisation would have to look like in future. In my view this is necessary. Otherwise it is too easy to get caught in existing processes and ways of thinking, and the status quo becomes the obstacle.

Of course, no one rebuilds an organisation from scratch. The transformation proceeds step by step, through individual activities, mandates or client segments. Once the goal is defined, each step must answer one question: does it move the organisation closer to the target architecture, merely support the existing workflow, or cement it?

What the board can take away

Those willing to run the thought experiment are likely to find new opportunities. It requires accepting one premise, however uncomfortable: highly standardised tasks and repetitive knowledge work will be automated in future. We therefore have to ask ourselves:

  • How would we build the organisation if we could start from scratch today, on a greenfield site?
  • How will our clients work in future?
  • What then remains our core task, and what do we take responsibility for?
  • How is our environment changing, and what opportunity does this create?

The gap between the ideal greenfield state and today’s situation defines the scope for strategic development and sets the direction. The answers form the starting point for the strategy discussion. The new vision calls for entrepreneurial courage. Without it, all that remains is the use of tools.

Transparency note

Generative AI was used in a supporting role for research and editorial preparation of this article. The conceptual design, the selection and weighting of the content, the professional assessment, the final wording of the text and the responsibility for all statements and conclusions rest with me.

References

This article was first published on the blog of the SwissBoardForum, the leading platform for boards of directors in Switzerland.