Microsoft Switzerland: AI Rollouts Need Workflow Redesign as Only 24% See Leadership Alignment

For IT leaders, CIOs and Microsoft 365 administrators, the useful part isn't the general case for AI. It's how Microsoft's own account of enterprise adoption has shifted over the past few years.

What Hinkel says, according to Microsoft's summary​

In Microsoft's summary, Hinkel says what matters is no longer access to technology alone, but the ability to connect and manage fragmented data, redesign workflows, and link innovation to clear business objectives.

Her main points:

• Put AI into core processes, not just side projects. Hinkel says AI delivers its greatest benefit when it is integrated into everyday work and core business processes.
• Don't simply automate the old way of working. According to the summary, companies should not simply automate existing processes, but rethink how people and AI work together.
• Move routine work off people's desks. Employees could then hand off routine tasks and focus more on creative, analytical, and value-adding activities.
• Keep humans in charge. The summary says human judgment, critical thinking, and clear governance remain central: 84 percent of Swiss AI users regard AI outputs as a starting point rather than a final answer.
• Meet the prerequisites. Trust, digital sovereignty, and skills are also important prerequisites for a successful transformation.
• Change leadership and culture. For adoption to produce measurable value, the summary says companies need not only a modern cloud and data foundation but, above all, clear leadership, a culture of learning, and a willingness to fundamentally reorganize work.

She also points to the high level of AI usage in Switzerland and sees great potential, particularly in the innovation landscape of French-speaking Switzerland (Romandie).

In short: this is a leadership interview, not a product announcement. It contains no new feature, licence change or deployment guide. Its value is as a statement of how Microsoft wants enterprise customers to think about AI over the next phase.

Where the 84% comes from​

The summary doesn't give a source for the 84% figure. A Microsoft Switzerland press release from July 9, 2026 does. The figure comes from Microsoft's 2026 Work Trend Index, which Microsoft says draws on trillions of anonymised Microsoft 365 productivity signals plus a survey of 20,000 AI users in 10 countries.

That release adds more Swiss numbers:

Microsoft defines Frontier Professionals as people in organisations that build AI into their workflows and redesign how work gets done. It says this group is nearly 20 percentage points more likely to report producing work they couldn't have done a year earlier.

Read the numbers with care​

• They are self-reported. The 84% shows how surveyed users say they treat AI output. It doesn't prove they check that output well, or that the output is reliable.
• Correlation isn't causation. The Frontier Professionals gap is an association. Organisations that redesign workflows may already differ in budget, skills or culture.
• The Swiss sample size isn't given. The July release doesn't say how many of the 20,000 respondents were in Switzerland, which makes the national figures hard to judge.
• Microsoft is the source. It sells the tools whose value the survey describes. That doesn't make the data wrong, but it's a reason to weigh it carefully.

The most revealing figure is also the least flattering one: only 24% say their leadership is aligned on AI. Many Swiss employees report personal gains, but most don't see a clear organisational strategy. That gap supports Hinkel's point that leadership, not software, is the bottleneck.

How Microsoft's message has changed​

Microsoft's own archive shows how Hinkel's public message has developed:

• January 2024, Davos. Speaking to L'Agefi, Hinkel said companies could achieve productivity gains of 20% to 60%, as shown by positive feedback from customers such as Axpo or Sunrise, who successfully use Microsoft Copilot. That message was about productivity, with named products and customers.
• January 2025, SRF Eco Talk. She argued that through AI solutions like Microsoft Copilot, SMEs can boost their efficiency and strengthen their competitive position, and stressed ensuring that humans remain in control while AI serves as a supporting tool.
• June 2025, Swiss Economic Forum. The focus was Microsoft's $400 million investment in expanding local cloud and AI infrastructure in Switzerland. Microsoft said this was meant to make modern technology easier for Swiss companies to access while maintaining the highest standards of data sovereignty and security.
• October 2026, Focus/Bilan. The new summary names no product and quotes no productivity percentage. It talks about workflow redesign, governance, leadership and learning culture.

Over two and a half years, the message has moved from "here are the gains" to "the gains depend on reorganising the business". That's a more modest pitch, and probably a more realistic one. It also shifts some responsibility for disappointing AI results onto customers' own organisations.

The 2025 investment also helps explain why digital sovereignty sits beside trust and skills on Hinkel's list. The October summary doesn't define what she means by sovereignty or say how it applies to specific cloud setups, so it would be speculation to fill that in.

What IT teams can take from this​

None of this is official guidance, but the summary and the survey data point to some practical lessons for administrators and IT leaders.

1. Fix the data first. Hinkel's mention of connecting and managing fragmented data is the practical core. AI assistants work with the content and permissions they can reach. Messy data and overshared permissions limit the results and increase the risk.
2. Measure redesigned processes, not licence counts. Usage figures show adoption. They don't show whether a process got better. Pick one or two core workflows, record a baseline, and then measure.
3. Make human review a written rule. The 84% figure describes what users say they do. Governance should make it policy: define which outputs need human sign-off and who owns them.
4. Train people to check output. Swiss respondents named quality control (46%) and critical thinking (42%) as critical skills. Training should cover how to verify AI output, not just how to write prompts.
5. Get leadership aligned before scaling. If three-quarters of users don't see a consistent strategy, rolling out more tools probably won't fix that.

In short: the interview gives no technical instructions, but its main argument, combined with Microsoft's survey data, offers a reasonable test for any AI rollout: are you reorganising work, or just adding licences?

Bottom line​

Hinkel's Focus interview restates themes Microsoft has been pushing all year: redesign work around AI, keep humans accountable, and treat data, skills, sovereignty and leadership as preconditions. The 84% figure comes from Microsoft's 2026 Work Trend Index. The same survey shows a weak spot: only 24% of Swiss AI users say their leadership is clearly aligned on AI.

For Windows and Microsoft 365 shops in Switzerland and elsewhere, the message is worth taking seriously, with one caveat. It comes from a vendor citing its own survey. The test is whether your own organisation, measuring its own workflows, sees the same results.