A mandate has a goal. We work toward that goal. It ends when the goal is achieved and the structures we have established are self-sustaining—not before.

ANKORA42 does not become a permanent part of an organization. Progress, leadership, and a focus on the desired outcome do not require an internal role. Those who commission our services receive our full attention, which is why we handle only a few projects at a time. The engagement ends when the organization is able to maintain the achieved level of performance on its own.

In most organizations that find themselves under pressure, it is not the will to act that is lacking. What is missing is a common ground on which decisions can be made. Reports contradict one another. Assumptions are treated as facts. Assessments are driven by vested interests.

A situation report consistently distinguishes between what is verifiable, what is an assumption, and what is an opinion. It is developed independently of existing program structures and their reporting logic. It is this independence that creates a solid foundation. The report also explicitly documents what is unknown: open questions, unchecked assumptions, and gaps in knowledge. This makes the situation report a basis for decision-making.

When measures are in place, decisions have been made, and yet there is still no effect. This situation is harder to recognize than an obvious crisis because, on the surface, the system still appears to be functioning.

This becomes apparent when decisions are effectively in force without having been formally announced or documented. When different assessments of the situation emerge between management levels and functional departments. When action replaces explicit decision-making. At this point, what is needed is not further internal analysis, but clarity regarding the cause, options for action, and binding implementation. We are convinced that this is where external objectivity and support are needed.

Through direct access to information, structures, and the people in charge. And through a strict separation between what the people—and the data behind these structures—report. A situation assessment distilled from existing reports merely reproduces their logic. It does not produce an independent picture.

In urgent situations, we provide an initial assessment within 72 hours. For more complex projects, this takes one to two weeks. The result is ready for decision-making: a concrete action plan that is robust and actionable. Not just another strategy presentation.

Because the real change isn’t happening in the applications themselves, but in the decision-making processes. In many organizations, AI initiatives are launched faster than responsibilities can be clarified. Decisions are delegated to systems whose behavior isn’t fully transparent. When responsibility isn’t clearly assigned, gaps in oversight arise.

This shift remains invisible at first. Processes function as they should, and results seem plausible. It is only when pressure mounts that it becomes clear that the logic behind decision-making no longer holds up. Those who treat AI as a transformation initiative gain speed at best. Often, projects are shelved. But those who treat AI as a management issue retain control.

Across five levels that do not function in isolation but as an integrated system: governance and decision-making logic, operational monitoring, architecture and integration, performance monitoring, and regulatory compliance. Each level on its own provides partial assurance. It is only through their interaction that a control center emerges that is sustainable over the long term.

In practical terms, this means that responsibilities are clarified before systems go live. Deviations are identified before they escalate. Compliance requirements such as the EU AI Act and ISO 42001 are integrated as a control mechanism, not as a brake applied after the fact.

Governance documents describe how decisions should be made. Decision-making architecture describes how they are actually made and where rights, principles, processes, and consequences diverge.

In well-functioning organizations, these four dimensions are implicitly present. Under pressure, they break down: responsibilities are informally shifted, fundamentals are not scrutinized, workarounds bypass formal structures, and consequences fail to materialize. The work on decision-making architecture is not aimed at creating new documents, but at achieving operational clarity: Who decides what, and on what basis?

Because they are invisible. Facts can be verified. Assumptions that are not identified as such cannot be verified and are passed on as established facts in subsequent decisions. Under pressure, this effect intensifies: speed takes the place of verification.

The consequence is not that bad decisions are made. It is that good decisions are based on a foundation that, upon closer inspection, is unsound. Distinguishing between facts, assumptions, and opinions is therefore not a mere methodological detail, but a prerequisite for sound decision-making.

The structures that were in place before the crisis did not prevent it. A fresh start therefore means rebuilding in a targeted manner, armed with the knowledge of what was missing.

In practical terms, this means that transitional roles are converted into defined responsibilities. Temporary processes that served as stopgap measures are put on a solid footing. The causes of the issues are documented and incorporated into the new structures. Operations are not handed over until these structures are self-sustaining.

The ability to maintain control is not lost all at once. The process is gradual and follows recognizable patterns: from a state of apparent stability, in which internal control mechanisms fail while the organization continues to function outwardly, to systemic escalation.

The signs are there. They are simply not being recognized as such. Activity is increasing without any real progress being made. Decisions are being postponed or only partially made. Priorities remain unclear, even though they have been formally defined. Anyone familiar with this pattern can determine the right moment to take effective action before it becomes necessary. The sooner action is taken, the less effort it will require.