Highlights
- Organisations often fail to realise AI’s promise because strategies overlook the human element in adoption and execution.
- Unlocking AI requires a structured framework that balances technology with usability and business impact.
- Lasting success comes from empowering people, shifting mindsets, and embedding AI into everyday ways of working.
AI promise vs. reality
The promise of artificial intelligence has captivated boardrooms across Europe. Companies are investing heavily, developing ambitious strategies, and envisioning a future transformed by intelligent systems. Yet, for many, the reality falls short of the aspiration. What I have observed over the past two years is a recurring bottleneck - the gap between AI strategies and their successful implementation. The data backs this up, with the European Commission noting that in January 2025 only 13.5% of enterprises in the EU were using artificial intelligence.
However, this isn't a technology problem. The real reason enterprise AI often fails to reach its full potential is not the sophistication of the algorithms or the power of the hardware but fundamentally ignoring the human element.
Too often, AI initiatives are launched with impressive ROI projections and complex technical designs but without a clear, actionable plan for engaging the very people who will use, manage, and benefit from these systems. Digital transformation, at its core, demands user buy-in and understanding. If the marketing team cannot seamlessly share customer data, or if the finance department does not grasp how AI can enhance forecasting, the entire endeavour will struggle.
Ultimately, the whole team should work together like an orchestra when implementing AI. With the technology specialists to the marketers and even the HR team acting as musicians creating AI-supported output, and the business leaders driving the implementation forward as the conductors.
Unlocking AI’s potential
Effective transformation requires a shift in approach, moving from a purely technology-centric view to one that is deeply human-centric from the very start. To truly unlock the potential of AI, businesses must begin by asking the right questions that place people at the heart of the equation. Every stakeholder, from the CEO to the front-line employee, should be able to articulate "what should AI do for my function?" This prompts the identification of specific, practical use cases that address real business challenges, demanding a deep understanding of daily realities and how AI can make work easier, more efficient, or more impactful.
Once what AI can do has been established, the next question should be "How can I enable AI use for my function?" This empowers stakeholders to take ownership, providing them with the resources, training, and support needed to access data, build agents, and integrate AI into their workflows.
Finally, - "How do I ensure AI is adopted?" This emphasizes creating a user-centric experience, making AI tools intuitive, valuable, and even enjoyable. Ultimately, it is about designing systems that complement, rather than replace, human capabilities, ensuring users understand what is in it for them and the part they play in the success of the team orchestra.
Implementing AI strategy
Bridging the gap between AI strategy and implementation demands a structured yet adaptable framework that makes the technical side of the implementation robust and long-lasting, and the ‘people’ side simple to roll out for employees. This framework must build a solid foundation of data, knowledge, and intelligence, while simultaneously prioritising usability for employees, security, and ethical considerations.
Here is an example of a multilayered approach we like at TCS that establishes a robust, observable, compliant, and scalable foundation, but also acknowledges the centricity of the human element to AI success:
- At its base is the data foundation. This involves establishing the right architecture, governance, and management practices to ensure enterprise data is clean, trustworthy, secure, and accessible.
- Building upon this is the knowledge layer. AI does not operate in a vacuum. It needs context. Enterprises should focus on capturing company processes and curating tacit organisational knowledge, making it accessible to AI systems.
- The intelligence layer is where the models begin to take shape. This involves leveraging Large Language Models and other advanced models to create agents that can utilise available data and knowledge to solve specific business problems.
- Next are the work systems, integrating AI into the day-to-day operations of the business. It involves building systems that are autonomous, semi-autonomous, augmentative, assistive, or transformative, following the "Describe, Predict, Diagnose, and Prescribe" philosophy. To make work easier, more efficient, and more impactful for employees.
- Crucially, guardrails must be embedded throughout the framework. AI must be deployed responsibly, with a strong emphasis on ethics, data protection, and regulatory compliance. These guardrails are essential for safe, ethical, and compliant AI use.
- Finally, usability brings the human-centric approach full circle. This involves creating AI systems that are user-friendly, intuitive, and valuable. It is about designing for multi-channel access, testing with focus groups to validate usability, and ensuring that users clearly understand the benefits they will gain.
For this framework to succeed, AI must be observable, compliant, and scalable. Observability means tracking AI performance and adoption through both use case-specific and general metrics like response times, uptime, sustainability, data volumes, and costs. Compliance demands adherence to all relevant regulations with built-in mechanisms for monitoring and auditing. Many organisations will include considerations for regulations like the EU AI Act as well to ensure longevity of the model.
Lastly, scalability requires a "perpetually adaptive" AI architecture that allows for flexible extensions and reusable components as technologies and workforces evolve.
Empowering people and shifting mindsets
It is important to remember that a framework is still a blueprint (no matter how robust). Success in AI implementation ultimately hinges on empowering people and orchestrating change at every level. The AI technology is the music and, as we have covered, our orchestral musicians are the diverse stakeholders across the organisation: the CMO, CFO, CHRO, and everyone in between, each with unique knowledge and skillsets. And just as a conductor guides musicians, leadership must provide a clear vision and ensure everyone is aligned. Without buy-in across the business, the entire AI strategy will falter.
Being committed to empowering people to embrace AI and drive innovation through a multi-pronged approach is essential and can be achieved in several ways. For example, at TCS, a step we have taken is to democratise agent creation, so that AI is not confined to a small group of data scientists. Rather, core teams of data scientists and AI architects are complemented by talent in business units who understand both their domain and the potential of AI.
In simple terms, turning AI strategy into action is building a strategy, developing a framework, and then most crucially planning human-centric implementation. By focusing on the human element of AI adoption, it is possible to drive tangible value for organisations and clients.
