By Jesper Lowgren, Chief Enterprise Architect, DXC Technology
AI is evolving in leaps and bounds. It has already transformed industries, and its impact will only grow.
But why is this such a big deal?
What does it mean for enterprises?
How can organizations prepare for an AI-driven future?
Let’s begin by putting this into perspective.
The purpose of any technology is scale—to do more, faster, and cheaper, or to enable entirely new possibilities.
Over time, we have outsourced almost all physical work, first to mechanical machines, later to computers. Put simply, we have been scaling our doing.
However, until recently, scaling has been limited to doing, not thinking or reasoning—which we have reserved exclusively for ourselves.
But that has changed.
Large Language Models (LLMs) and autonomous AI systems are now capable of high-level reasoning and decision-making—at a fraction of the cost of human cognitive labor.
This warrants a moment’s reflection:
Human-level reasoning at scale, at a fraction of human cost.
The potential is staggering, and just like the Gold Rush and the Internet Boom, the AI rush has begun. However, the difference this time is the insane speed of evolution and the unprecedented risks for organizations that are unprepared.
The Challenge: Managing Loss of Control
AI autonomy introduces massive opportunities—but also massive risks. The issue isn’t just AI’s capabilities; it’s how enterprises govern and manage the loss of control in a world where AI agents make independent decisions.
Today’s business processes, governance models, and enterprise architectures are not designed for autonomous AI agents. Organizations must rethink foundational principles and develop new frameworks, new governance models, and new ways of working.
While much remains to be defined, several key themes are emerging:
✅ More governance, less management: Transitioning from rigid, top-down control to adaptive governance that enables AI autonomy while maintaining oversight.
✅ AI-human collaboration:– Establishing trust-based systems where AI and humans work seamlessly together.
✅ Accountability and transparency: Ensuring AI-driven decisions are traceable, explainable, and aligned with ethical principles.
✅ Ethical considerations: Addressing bias, fairness, and unintended consequences, ensuring AI supports both innovation and responsibility.
However, autonomy without structure leads to disorder. This is where enterprise architecture (EA) becomes mission-critical.
Enterprise Architecture: The Foundation for Scalable AI
Enterprise architecture (EA) is no longer just a strategic asset—it is the cornerstone of AI scalability and governance.
Without a strong architectural foundation:
❌ AI systems will become fragmented.
❌ AI agents will lack interoperability.
❌ Enterprises will lose control over AI-driven decision-making.
To avoid these pitfalls, EA must evolve alongside AI. This transformation is best understood through a five-level maturity model, illustrating AI’s journey from a simple tool to a fully autonomous force.
CMMM Agentic AI Maturity Levels
The Five Maturity Levels of Agentic AI
Level 1: AI as a Tool
At this foundational level, AI is deterministic and task-based, performing well-defined functions such as automation, data analysis, and basic chatbots.
Enterprise Architecture Focus:
AI at this level exists in isolated, function-specific silos, requiring minimal structural adjustments.
- Focus is on achieving compatibility with existing technology stacks rather than driving transformational change.
- Data governance remains centralized, and business architecture continues to rely on linear, human-driven workflows.
Level 2: AI as an Assistant
AI enhances human decision-making, offering intelligent recommendations in areas like financial forecasting, risk assessment, and supply chain optimization.
Enterprise Architecture Focus:
To advance to this level, enterprises must transition toward a more integrated technology architecture.
- Data-sharing capabilities need to extend across functions, requiring interoperable APIs, federated data governance, and enhanced cross-departmental collaboration.
- Business processes shift from rigid, rule-based models to flexible, insight-driven operations that leverage AI to drive efficiency and effectiveness.
Level 3: AI as an Operator
AI begins to execute business operations independently, orchestrating workflows, optimizing supply chains, and managing dynamic pricing.
Enterprise Architecture Focus:
EA at this stage must evolve to support real-time data flows and scalable governance frameworks.
- Unified data architectures become critical, allowing AI to access and act upon high-quality, real-time data across the enterprise.
- Cloud-native solutions, containerization, and microservices architectures become essential to ensure agility and scalability.
- Governance structures must be updated to incorporate ethical AI guidelines, compliance measures, and bias mitigation strategies.
Level 4: AI as an Actor
At this stage, AI actively shapes business outcomes and collaborates strategically with humans, rather than just executing predefined operations.
Enterprise Architecture Focus:
The evolution to this level requires a paradigm shift in enterprise architecture.
- Agile, event-driven architectures become the norm, enabling organizations to respond dynamically to market conditions.
- AI governance moves from static rules to adaptive frameworks, ensuring AI operates ethically and transparently.
- Organizational structures shift toward AI-human hybrid models.
Level 5: AI as an Autonomous Force
AI operates with near-complete autonomy, driving real-time business strategy and innovation at scale.
Enterprise Architecture Focus:
Enterprise architecture in this paradigm must be fully dynamic, ecosystem-centric, and modular.
- AI-first architectures emphasize decentralized decision-making, distributed AI governance, and cross-organization AI Agent transactions.
- Technology platforms transition to real-time, serverless computing models, leveraging edge AI and quantum computing for ultra-fast processing.
Architecting Scalable Foundations for Agentic AI
The rise of autonomous AI and AI Agents is not merely a technological evolution—it represents a fundamental shift in how businesses operate, innovate, and compete. For the first time, we are extending the boundaries of automation from physical tasks to complex reasoning and decision-making, enabling organizations to scale thinking, creativity, and problem-solving. However, this leap comes with significant challenges, especially for organizations navigating this transformation without a clear architectural foundation.
Enterprise architecture is no longer just a strategic asset; it is the cornerstone of scalability in the era of Agentic AI. It provides the structure needed to align AI capabilities with business goals, ensuring seamless integration across people, processes, data, and technology. Without a robust enterprise architecture, organizations risk creating fragmented systems, disconnected AI agents, and unintended complexity—undermining the very benefits AI promises to deliver.
As AI reshapes industries, enterprise architects become natural stewards of transformation. They must design scalable frameworks that enable AI agents to interact not only within the organization but across ecosystems, ensuring compliance, interoperability, and trust. This isn’t just about managing technology—it’s about building a foundation for sustainable innovation that empowers businesses to harness AI’s full potential.
The future of Agentic AI demands more than technical advancements; it requires a cohesive vision rooted in architectural principles. By embracing enterprise architecture as the blueprint for this transformation, organizations can ensure that AI serves as a force for progress—scaling not only efficiency and output but also resilience, adaptability, and ethical responsibility.