Local AI: Governance Before Hardware Purchase
Benedikt Langer
10 min readFour developments over two weeks show that locally operated AI goes far beyond the tech stack. ...
Read Article
3 min read
TL;DR
- 70% of all transformation projects miss their goals – the most common cause is organizational resistance, not technology.
- AI initiatives fail especially often due to lack of adoption: employees distrust AI recommendations and revert to manual processes.
- The 30-70 Rule: 30% technology, 70% organizational change.
- Three levers for adoption: psychological safety, tangible benefit, and progressive autonomy.
- The CEO must make change a top-priority leadership issue – delegating it solely to HR guarantees mediocrity.
An insurance conglomerate invested €15 million in AI-powered claims processing. The technology delivered: 40% faster processing, 25% fewer errors. After six months, only 12% of claims handlers used the system voluntarily.
The technology isn’t the problem. The organization is. Change management determines whether that €15 million investment becomes a strategic breakthrough – or a €15 million misstep.
Why AI Triggers Unusual Resistance
AI threatens three fundamental human needs: competence (the AI accomplishes in seconds what took me years to master), control (opaque, “black-box” recommendations without transparent reasoning), and job security (public discourse centers on job losses).
Passive non-adoption is enough to sink an AI project. Employees don’t need to sabotage – it’s sufficient that they devise workarounds.
The 30-70 Rule
In successful transformation projects: 30% technology, 70% organizational change. Reality? 90% of the budget flows into technology; just 10% into change.
Communication: Not one-off launch events – but ongoing, honest updates.
Training: Hands-on, supported practice over weeks – not two hours of e-learning. Champions embedded in every team.
Process redesign: Rebuild workflows around AI’s strengths – not force AI into legacy processes.
Three Levers for Adoption
Psychological safety: Making mistakes while learning carries no negative consequences. Leaders experiment themselves – and openly discuss their own learning curves.
Tangible benefit: When AI takes over hated documentation tasks, acceptance for everything else rises sharply.
Progressive autonomy: First, AI recommends; humans decide. Teams set their own pace for increasing reliance.
The CEO’s Responsibility
Set the narrative: Communicate honestly what will change – and what won’t.
Hold leaders accountable: Middle management decides success or failure. Define AI adoption as a core leadership responsibility.
Protect resources: The change budget is always the first to be cut. The CEO must defend it.
Frequently Asked Questions
How much budget should go to change management?
30-40% of the total project budget. For €1 million in technology costs: €400,000-€600,000 for change.
How do I measure success?
Adoption rate (% of active users), usage intensity (basic vs. full functionality), and employee satisfaction. Under 60% adoption after six months signals a change deficit.
What if leaders themselves resist?
One-on-one conversations, peer pressure from colleagues who’ve succeeded, and tying AI transformation goals – including bonus relevance – to performance agreements.
Header Image Source: Unsplash / Jason Goodman
Read more
- Gartner Technology Trends: 2026 – a year of disruption, innovation, and risk
- Germany’s internet economy grows by nearly 10% annually
Explore more: Additional articles on mybusinessfuture