By Gabriel von Mitschke-Collande, Group Chief Digital Officer (CDO), Giesecke+Devrien

AI has moved beyond the stage of being a technology organizations can simply observe from the sidelines. For many businesses, the question is no longer whether AI will influence how work gets done, but how quickly and responsibly it can be integrated into products, processes, decision-making and organizational structures.

For organizations operating in security-critical domains, however, adoption comes with a fundamental requirement: trust.

Our relationship with technology is changing. Until now, humans have had to learn the language of machines through coding, commands, interfaces, and processes. AI is changing that relationship, with machines increasingly learning to communicate in our language. In other words, we enter in an era of “humanization” of technology.

AI can interact with us, answer questions, make recommendations, automate workflows and increasingly act on our behalf. That is more than a change in interface. It changes the human-technology relationship, making trust and values even more important. As technology takes on a more active role, expectations surrounding transparency, security and accountability inevitably increase.

Trust, therefore, is not simply an ethical consideration, it is a prerequisite for meaningful adoption. If AI can communicate, recommend, decide and act for us, trust cannot be an afterthought. It must be built into how the technology is designed and used. More than that, trust is a business enabler: it is what gives people and organizations the confidence to adopt AI, integrate it into their operations and ultimately scale its use.

Trust starts with transparency

AI is an enormous and rapidly evolving field, and organizations should resist communicating about it in vague or overly ambitious terms. Employees, customers and other stakeholders need to understand what an organization is doing with AI, why it is doing it, what is known and where uncertainty remains. Only under those circumstances can trust be built.

This is particularly important because modernization creates questions as well as opportunities. Employees may wonder how AI will affect their roles, while customers may want to understand how their data is being used.

There is no value in pretending every answer is already known. AI will continue to evolve, and organizations will discover new applications and challenges along the way. What matters is establishing a clear direction and communicating it consistently.

Move beyond experimentation

The early stages of AI adoption often focus on individual use cases and experimentation. That phase is valuable for understanding what the technology can and cannot do. But eventually, organizations need to shift from asking what AI can accomplish to asking where it can create measurable, repeatablevalue and how business models will changed based on AI. This is not only important for digital products, but also relevant for physical ones.

That value can take several forms: improving productivity, reducing costs, accelerating processes, improving time to market or creating new products and services.

The distinction between experimentation and transformation matters. Saving five minutes with AI is useful, but the greater opportunity comes when AI becomes part of the workflow, automating repetitive steps across a process. The real value emerges when AI moves beyond isolated tasks and becomes embedded in how work is actually done. That is where organizations can create scale, generate meaningful efficiency gains and build momentum around adoption. The critical changeis bringing technology and business perspectives together. AI cannot remain isolated within a technology function. Organizations need people who understand both the technology and the specific problems it is intended to solve.

Build the foundation before scaling

There is considerable attention around what AI can accomplish, but less attention is paid to the foundational work required to make those capabilities work effectively at scale.

Data is one example. Organizations need to understand what data they have, how it is structured, who can access it and whether it is appropriate for a particular AI application. They also need an IT architecture capable of connecting the systems and information required.

Security is equally important. Organizations handling sensitive information cannot simply introduce AI without considering where data goes, how it is protected and what safeguards are required. In payments, digital identities and critical infrastructure, trust is not optional. If people cannot trust the systems behind these services, adoption simply does not happen. This is particularly important from a securitytech perspective, where security, privacy and resilience are fundamental to the services and systems people rely on.

This is where AI and organizational design intersect. Before AI can automate repetitive work, organizations need to understand, document and often simplify existing processes. AI does not replace process discipline; it makes it more important.

Be digital, stay human

Technology alone does not create an AI-ready organization. People remain at the center of the transformation.

“Be digital, stay human” is a useful principle for responsible AI adoption. The goal is not simply more technology, but using it to create value while preserving human judgment, accountability and values. AI will modernize how people work, making it essential to consider the human side of that transformation. Employees need access to appropriate tools, opportunities to experiment and training that helps them understand both AI’s potential and its limitations.

AI literacy should become an organizational capability. People need enough understanding to work confidently with AI, recognize its limitations and know when its output requires human review.

Leadership has an important role, too. When leaders openly demonstrate how they use AI, they can encourage employees to experiment, learn and develop their own capabilities. Governance should also extend into individual business functions, bringing together AI expertise and domain knowledge.

Ultimately, responsible AI is a value-driven discipline. Technology may provide the capabilities, but people determine how those capabilities are applied. That also means considering our broader societal responsibility. As AI becomes more deeply embedded in everyday life, we have to think not only about what we can build today, but about the kind of technology-enabled society we want to create for future generations. Innovation and human values should advance together, not compete with one another.

Applied AI requires an ecosystem

Scaling AI also requires access to talent and collaboration beyond the boundaries of a single organization. Strong AI ecosystems can connect research institutions, businesses and startups, helping translate research into practical applications. AI literally overcomes the limits of time and space, accessing knowledge distributed around the globe in real time.

That transition, from research to commercialization and real-world use, is increasingly important. The objective should not be AI for its own sake, but applied AI that solves meaningful problems while maintaining appropriate standards for security, privacy and accountability.

Ultimately, scaling AI is not about finding a single technology or application that changes everything. It is about building the capabilities to use AI responsibly and effectively over time.

Organizations need strategy, governance, secure infrastructure, reliable data, well-designed processes, skilled people and leaders willing to communicate honestly about both progress and uncertainty. More importantly, they need a clear understanding of the value they are trying to create and the human principles that should guide the technology along the way.

AI may be advancing rapidly, but successful adoption will depend on something more fundamental: whether people trust the systems being built.

The more AI changes the relationship between humans and technology, the more important that trust becomes. In the AI age, that trust will increasingly be built on a combination of human judgment and technological analysis. AI can uncover patterns and insights that may go beyond what an individual human can see – for example, when analyzing a patient’s blood results. But trusting AI does not mean replacing human judgment. In security-critical environments, trust cannot be added after deployment. It has to be designed into the AI journey from the beginning.

About the Author: Gabriel von Mitschke-Collande is the Group Chief Digital Officer (CDO) of Giesecke+Devrient, a global SecurityTech company. He is responsible for the strategic direction of data and AI, as well as, and drives the development of forward-looking business models and solutions. For more information, please visit www.gi-de.com.