Topleadership through AI disruptions - 3

This is part 3 of three on top leadership of AI disruptions

AI is more than merely a new technology. It is a universal catalyst that enhances human functions, cognitive, communicative and creative, for all individuals willing to use it. Compared with previous technological disruptions, AI has a far more intrusive potential, because it fundamentally alters the conditions for leadership, judgement, capabilities and competition. For this reason, AI adoption is also most effective when it happens bottom-up.

The task of top management is therefore primarily to encourage and motivate the use of AI, within carefully considered security frameworks.

Therefore, success with AI ultimately depends on top management’s mindset and motivation …

At a fundamental level, AI is inherently risky because it confers power at a speed that far exceeds the pace at which wisdom can be developed.

AI therefore compels top leaders to rethink the classical tension between operations and renewal.
It is no longer merely a question of executing efficiently, but of building organisations that are able to learn (almost) as fast as the technology itself. In the end, it is not AI that changes companies; it is employees who determine how AI will best change them. As noted above, the leader’s role is to stimulate curiosity and to set boundaries.

Taken together, the AI winners of the coming years will therefore be those who are able to:

• combine ethical grounding with technological courage,
• create trust throughout the transition (which starts with trust within the organisation), and
• translate technological potential into human meaning.

… where the requirements vary depending on how far the AI impulse has progressed

The requirements can be illustrated using the structure from the previous series of blog posts:

1. Before AI‑intensive operations (the past three years and at most the coming year)
   1. Establish data discipline, in particular a shared understanding of what data is and what its effects are
   2. Map current technological dependencies (risk management).
   3. Build organisational understanding of AI’s possibilities and limitations (for example by creating “sandboxes” and assigning objectives to them)
   4. Identify where differences in capabilities and motivation exist
2. During AI acceleration
   1. Create clear priorities: where can we already see that AI delivers real value? Where is the risk–reward balance favourable, and where do we realistically have opportunities to position ourselves?
   2. Calibrate frequently: use small loops rather than large, pre‑defined projects
   3. Involve experts who understand both the domain and the technology
   4. Manage the pace, not the technology. Too high or too low a pace will undermine organisational motivation
3. After the first AI wave (the learning environment)
   1. Consolidate data flows, roles and processes
   2. Re‑establish normal operations and evaluate outcomes
   3. Build and embed “AI real capital”, i.e. accumulated experience, practices and human judgement

From a governance perspective, this is about making complexity simple and engaging …

Managerially, this means:

• Separate streams. Distinguish between production AI and experimental AI. Avoid running the entire organisation in beta mode; this easily confuses and frightens stable operations. “Sandboxes” enable learning without blame or fault. Ethical safety is a prerequisite for technological courage. Establish a small, stable and decision‑capable team in which the CEO participates, but does not sit at the head of the table. Avoid large steering committees. Consistent execution is critical. There are no central projects, but rather a multitude of decentralised tools that must be adopted as effectively as possible. In AI transformation, continuity matters more than speed.
• MeasureThe value of AI must be measurable in terms of quality, time or risk, not in PowerPoint. AI’s results are often invisible at the outset (efficiency gains that are offset by the time spent evaluating and understanding them).
  • Communicate concrete outcomes, such as: “We reduced customer response time by 40%.” 
  • Signal a clear ethical direction, for example: “No AI decisions without human oversight.” 
  • Be wary with KPIs and traditional bonus systems. AI needs to be played into the organisation in order to create ownership. 
• Engage. Define the purpose. Start with the desired outcomes, not with the technology.
  • Build a simple narrative for AI based on this. A sense of urgency can be created by demonstrating how far competitors are already moving with AI, but motivation must ultimately be positive and internally driven, via opportunity, not fear. 
  • Democratise ideas, but centralise direction. AI affects everyone (“universally enabling”), but the learning process must be coordinated so that the company’s products and services remain aligned with its brand and credibility.
• Be realistic. Do not overstate AI’s immediate impact, and do not underestimate its profound long‑term effects. Communicate that AI is a tool, not a replacement. Rumours and fear can erode trust if leadership is not proactive, including in communication. Think ahead and anticipate the next wave, especially what it implies for equipment, people and go‑to‑market. Avoid moving too early.

… and this must be supported through leadership practice

• Create possibilities. AI should not replace judgement; it should liberate it. “No” is never an answer, it should be the conclusion of careful consideration.
• Honesty. Be honest about AI’s weaknesses and strengths, but above all about where the organisation can improve its ability to capture opportunities.
• Openness. Limit prohibitions and instead govern access, data and purpose. Execution does not take place in rigid systems that behave identically every time. Execution is the product of individuals’ daily efforts. This is why self‑learning mechanisms are particularly important.
• Trust. Ask employees to try, not to perfect. Innovation requires repetition and courage. Follow up on initiatives so that employees and managers feel that what they are doing matters.
• Acknowledge. AI generates errors; they are valuable when discovered early. Celebrate mistakes, especially your own. This is a sign of honesty, openness and trust.
• Involve. Establish feedback cycles in which employees themselves demonstrate what works in practice. Remember to acknowledge successes. Ask your closest customers for their views on AI‑enabled services.
• Empower. Delegate responsibility for AI use early, but within clear boundaries. Playfulness is a prerequisite for creativity.

In conclusion, AI is a bottom‑up disruption that affects universally

AI differs from most disruptive impulses by having both broad and deep impact. The technology must therefore emerge decentrally and be integrated with the organisation’s existing functions. It affects all employees, and in ways that no one can yet fully anticipate. 

It is therefore the responsibility of top management to encourage and motivate the use of AI, within carefully considered frameworks for security and governance.