AI and sensitive sectors: taking the time to do it right
In this age of social networking shorts and unbridled marketing expectations, we at Darest have chosen to slow down a little. According to Brandolini's law, "the amount of energy needed to refute nonsense is an order of magnitude greater than that needed to produce it". So let's take the risk of investing this energy in a calm discussion of what AI will really be like in September 2025.
Why this thoughtful approach?
Our specialization in the Finance, Healthcare and Education sectors gives us a particular responsibility. These sensitive, complex and highly regulated fields of activity cannot afford the hasty adoption of AI. When it comes to financial data, medical information or educational pathways, error is not an option.
This implicit responsibility guides our approach: between the blind enthusiasm of technological evangelists and the sterile resistance of technophobes, there is a space for critical reflection that we must invest.
The Swiss AI paradox
The figures are striking: 52% of Swiss companies say they have deployed AI on a large scale (+31 points in one year), but only 35% have truly integrated it into their strategy. This paradox reveals an "illusion of understanding": we believe we have mastered AI because we know how to use it, without grasping its systemic implications.
In our sectors of intervention, this illusion can have dramatic consequences. Misunderstood AI in healthcare can compromise diagnoses. In finance, it can amplify discriminatory biases. In education, it can deepen inequalities.
The Swiss advantage: "reasoned" adoption
The Swiss ecosystem has some remarkable specificities for our sectors:
- A culture of precision (essential in healthcare)
- Commitment to data quality (62% of organizations rate it as "good to excellent")
- A historical ability to integrate innovation into stable institutional frameworks
These assets are all levers for the responsible adoption of AI in sensitive areas.
New critical territories
AI is spreading beyond marketing (77% of current uses) to finance, legal and supply chain. This expansion reveals a logic of "augmentation" rather than substitution.
Autonomous agents - systems capable of planning and chaining actions - mark a particularly critical qualitative breakthrough in our sectors. An autonomous healthcare agent that makes decisions without human supervision raises major ethical and regulatory issues.
Our approach: three principles for sensitive sectors
- Enhanced strategic intentionality
In our fields, every AI deployment must not only be part of a clear vision, but also comply with specific regulatory frameworks (FINMA, healthcare data protection laws, educational regulations). - Reversibility as an imperative
In finance, healthcare and education, maintaining back-up human skills is not conservatism - it's a deontological and regulatory obligation. - Sectoral collective learning
Our fields require the development of specific metacognitive skills: how do you audit a medical AI? How to explain an algorithmic financial decision? How can we guarantee educational equity in an AI-enhanced system?
The art of responsible temporality
Our sector specialization imposes a different timeframe on us. Where a start-up can "move fast and break things", we have to "move smart and fix things". Because in our fields, broken things mean lives, savings and learning paths.
AI is not our destiny - it's our opportunity. But for finance, healthcare and education, this opportunity can only be seized with the rigor, ethics and critical intelligence that characterize these sectors.
At Darest, we're convinced that the future belongs to organizations that know how to combine technological excellence and industry responsibility.