TLDR: As agentic AI emerges as a transformative force, projected to generate hundreds of billions in economic value by 2028, technology leaders like Microsoft and Capgemini are confronting significant challenges in its control, integration, and effective deployment. Industry experts highlight the critical need for robust data infrastructure, workflow-embedded analytics, cloud migration, and hypothesis-driven experimentation to truly harness the power of these autonomous AI systems, moving beyond mere impressive demonstrations to impactful business solutions.
The advent of agentic AI, autonomous systems capable of planning, decision-making, and adapting independently, marks a pivotal shift in the artificial intelligence landscape. This transformative technology, which can respond to plain language prompts and execute complex tasks without rigid rules, is poised for explosive growth and significant economic impact. Industry analysts like Gartner have already identified agentic AI as a top 25 technology trend for 2025, with IDC projecting a tenfold expansion to over $4 billion by 2028 in this category alone. The broader agentic automation market is expected to triple from approximately $300 million to $900 million between 2024 and 2025, maintaining a robust 36% compound annual growth rate through 2028.
Executives are keenly aware of this potential, with 71% anticipating that AI agents will drive higher automation in their workflows, and 52% planning implementations in 2025. A striking 93% of leaders believe that organizations successfully scaling AI agents within the next 12 months will gain a decisive competitive advantage. Projections indicate that AI agents could generate up to $450 billion in economic value through revenue uplift and cost savings across surveyed countries by 2028, with a potential to unlock $3.6 trillion by 2028 if all organizations achieve anticipated benefits.
However, this rapid adoption is not without its complexities and challenges, which major players like Microsoft and Capgemini are actively addressing. A recent Forbes Technology Council report, published on August 11, 2025, highlighted that while AI dominates corporate agendas, many companies are overlooking foundational technologies and strategies crucial for innovation, security, and scalability. This oversight can hinder the true impact of AI initiatives.
Experts emphasize several critical areas being overlooked. Mike Capone of Qlik points out a fundamental flaw in many current deployments: ‘Agentic AI sounds powerful, but most deployments are just assistants without context. If agents can’t access enterprise workflows, governed data and real-time signals, they can’t plan or act. They can only chat. The overlooked tool? Workflow-embedded analytics. Without it, AI stays impressive, not impactful.’ This underscores the necessity of integrating AI agents deeply within existing operational frameworks and providing them with actionable data.
Cloud migration is another foundational step often bypassed. Martin Taylor of Content Guru warns, ‘Too many companies chase AI without first moving to the cloud, which is vital for AI success. It’s similar to installing a jet engine in a car—AI can only optimize a system that is already built to handle its data needs, which legacy, on-premises solutions cannot do.’ The scalability and data processing capabilities of cloud infrastructure are paramount for effective AI deployment.
Anil Pantangi of Capgemini America Inc. stresses the importance of strategic methodology: ‘Everyone’s racing to implement AI, but many are running in the wrong direction. The most overlooked high-impact trend isn’t a new tool, but a classic method: rapid, hypothesis-driven experimentation. It’s the compass that guides your AI strategy, ensuring your powerful new engine is actually moving you toward customer value and business success.’ This highlights the need for a structured, iterative approach to AI integration.
Furthermore, Ashutosh Synghal of Midcentury Labs Inc. underscores the often-neglected aspect of data infrastructure: ‘Everyone’s chasing new AI models, but many ignore data tools—secure sharing, federated learning and clean data pipelines. Without that plumbing, the flashiest model stalls. Nail how you collect data first; the AI wins will follow.’ The quality and accessibility of data are foundational to any successful AI endeavor.
Also Read:
- Agentic AI Adoption Reaches One-Third Amidst Significant Integration and Governance Hurdles
- Agentic AI Adoption in Marketing Remains Nascent, Poised for Significant Growth
Capgemini, a key player in this space, has extensively researched the ‘Rise of Agentic AI,’ emphasizing the need for robust infrastructures that enable multiple agents to work across fragmented technology landscapes. They advocate for ‘orchestration’ as a critical capability for coordinating agents’ tasks, managing workflows, and facilitating collaboration between AI agents, RPA robots, and human workers. As organizations like Microsoft continue to push the boundaries of AI with tools like Copilot, the collective challenge for the industry remains not just in developing powerful AI, but in mastering its control, integration, and ensuring its real-world impact through meticulous planning and foundational technological readiness.
