How to Successfully Implement AI in a People-Centric and Sustainable Way Using Agile Principles

With Katja Koch, Senior Manager People & Culture at the management and IT consultancy MHP

A high percentage of AI initiatives fail. Not because of the technology itself, but because implementation efforts often focus solely on technical aspects. AI is not an IT project that can simply be “installed.” It is a highly complex transformation initiative. Roles, responsibilities, and ways of working are changing, requiring employees to demonstrate a high degree of flexibility and adaptability.

This is especially true for leaders. AI is no longer just a tool; it is increasingly becoming an actor in decision-making processes, particularly through the rise of AI agents. AI analyzes information and supports decision-making based on data. Today, humans still retain ultimate responsibility for decisions, such as prioritizing topics and validating AI-generated outputs (the “human in the loop” principle). In the future, AI systems will perform a growing number of tasks autonomously. As a result, leaders must learn to rethink and delegate responsibility. In many cases, this means less control and greater trust.

Closely linked to this is the emotional dimension. Many employees are concerned about their future and approach AI with uncertainty. Companies must therefore make change transparent and create opportunities for participation, not only for the management but for the entire workforce. Transformation can only succeed when people understand why it matters, what the organization aims to achieve with AI, and how they can actively contribute to shaping the change.

Prerequisites for Successful AI Adoption:

● A clear understanding of the business value AI is expected to create. Without this clarity, organizations risk ending up with a jungle of disconnected AI use cases that deliver little impact.
● An organization whose structures, processes, and culture are ready for AI. Otherwise, AI solutions may not be accepted or adopted.
● Agile principles, such as regularly reflecting on and improving AI solutions, gathering feedback early, and delivering usable results quickly to internal and external customers.
● A willingness to experiment, starting small and learning effectively how initial AI use cases can evolve into sustainable, enterprise-wide AI adoption.

Measuring Organizational Maturity

Before introducing AI, organizations should assess their level of organizational maturity. In other words, they need to understand where they currently stand. Many companies want to leverage AI but do not know whether their structures, processes, data landscape, or workforce are sufficiently prepared. A maturity assessment creates this clarity. It reveals which capabilities already exist, which are missing, and what is realistically achievable. Only with this understanding can organizations design AI adoption initiatives that move beyond the pilot phase and unlock their full potential.

In this context, leaders are particularly challenged to act with empathy and a strong focus on outcomes. The following leadership roles and responsibilities are becoming increasingly important: 

• Change Manager: Driving transformation by creating awareness, communicating the vision, addressing concerns, and actively guiding change
• Learning Leader: Leading by example, experimenting, and reflecting on their own mindset and behavior
• Collaboration Designer: Orchestrating collaboration between humans and AI, developing AI strategies, and redesigning organizational structures
• Enabler of Courage and Innovation: Encouraging creative thinking, challenging existing business models, and exploring new ideas
• Talent Developer: Developing both people and AI agents by evaluating AI outcomes and maintaining high decision quality

Another emerging trend is the need for teams to organize more strongly around products, solutions, or customer groups. This changes how success is measured: by outcomes rather than by the size of an individual department.

These organizational and leadership dimensions are often overlooked. Many companies invest too little time in building expertise, enabling practical AI usage in day-to-day work, and establishing the structural foundations required for lasting success.

Building the Right Organizational Structures

Organizations that understand their level of maturity can build structures that support successful AI adoption. Ideally, these include seamless data exchange, cross-functional teams, and strong AI governance that establishes clear guardrails and assesses potential risks.

At the same time, a closer integration of HR and IT is emerging. Some organizations already bring both functions together under a single leadership structure. This can be advantageous when introducing AI because technological decisions can be aligned early with questions about future roles and required skills.

As a result, the need for a clear enablement strategy comes into focus. Such a strategy defines the competencies required for different roles and outlines how employees and leaders will be prepared to work with AI. Training alone is not enough. Equally important is creating an environment where continuous learning and experimentation are part of everyday work. Proven approaches include Communities of Practice, where employees can exchange experiences and learn from one another.

Most organizations and their employees struggle to adapt quickly enough to the pace of technological change. Yet speed is the decisive competitive advantage. Companies that move too slowly risk falling behind. This need for speed is particularly relevant for leaders when making decisions. However, it does not mean everyone must simply work faster. Speed comes primarily from empowerment. The foundation of empowerment is a supportive leadership mindset. Leaders must enable employees to take ownership, ensure role clarity, and be willing to let others make decisions instead of retaining full control themselves.

Empowering Leaders

Leaders must not only steer transformation; they also need support in managing their own role transition. This includes trust, psychological safety, and traditional leadership development focused on strengthening leadership capabilities. Core competencies include: Systems thinking, Designing human-machine collaboration, Fundamental understanding of AI, data, and platform architectures, Knowledge of governance, ethics, and regulations (such as the EU AI Act).

At the same time, leaders must learn how to cope with the responsibilities placed upon them. Amid the complexity and uncertainty of transformation, it is more important than ever to pause, reflect, and ask critical questions: What are my own needs and preferred behaviors? Does AI strengthen them, or do I feel personally threatened by this new technology? Could I be unconsciously resisting change?

Agile Ways of Working Facilitate AI Adoption

Experience shows that organizations with an agile mindset generally find it easier to adopt AI. They are already familiar with the type of complexity that AI introduces and often possess the right mindset to handle it. Agile organizations do not think in rigid processes; they think in adaptable solutions. Agile teams tend to approach challenges more holistically and systemically, demonstrate greater courage to experiment, and work in a more iterative and structured way. They see not just the tools, but the bigger picture.
This open, learning-oriented mindset makes it significantly easier to embrace AI. To connect AI initiatives directly to strategic business objectives, the agile framework Objectives and Key Results (OKRs) can serve as an effective tool. Agile leadership principles and practices, such as “Take an Economic View” and systems thinking from the Scaled Agile Framework (SAFe), can also provide valuable guidance in navigating AI-driven transformation.