Agentic AI: Unleashing Potential, Navigating the Practicalities
Agentic AI promises transformative potential for enterprises, but its implementation faces key hurdles. While Capgemini Research Institute estimates its economic impact could reach US$450 billion by 2028, adoption remains surprisingly low—stuck at a mere 2% of organizations. This gap between hype and harnessing the technology highlights a fundamental issue: building trust and ensuring effective oversight.
Human-AI Collaboration Key for Success: New research from Capgemini, based on a survey of 1,500 executives globally, underscores the crucial role of human involvement in agentic AI workflows. A resounding 73% of respondents recognized the value of human intervention in AI processes. Importantly, nearly all (90%) perceived oversight as either beneficial or neutral in cost, reinforcing the need for human-centric AI implementation. Agentic AI, like a powerful tool, functions best when paired with human expertise and guidance, not left on autopilot.
Intent vs. Readiness: A Critical Gap: Current adoption rates paint a picture of a significant implementation gap. While a quarter of organizations have piloted agentic AI, and 14% have moved into implementation, the majority are still in the planning stages. This “intent-readiness gap” emerges as a major barrier to unlocking the promised economic value.
Real-World Applications Emerging: The concept isn’t theoretical. A personal shopping assistant capable of searching, describing products, answering questions, and adding items to a cart through voice or text demonstrates the practical application of agentic AI. While financial transactions may be excluded due to security concerns, this technology powerfully replicates human assistant functions, potentially reshaping online commerce. This raises the question: Could intelligent agents become the new face of online shopping, streamlining complexity and making eCommerce more accessible?
Defining Agentic AI: Beyond Generative Capabilities: To understand this emerging technology, we spoke with Jason Hardy, CTO for Artificial Intelligence at Hitachi Vantara. Hardy defines agentic AI as software capable of independent decision-making, action, and strategy refinement. Unlike generative AI, which creates content in response to prompts, agentic AI acts autonomously in dynamic environments, potentially utilizing generative outputs to achieve predefined objectives.
Driving Adoption: Scale and Complexity: Hardy attributes the growing interest in agentic AI to the sheer scale and complexity of modern enterprise challenges. Agentic AI offers more than simply analysis; it optimizes storage, automates compliance, anticipates failures, and responds to security threats in real-time, shifting from “insight” to “autonomous action.”
Value Proposition: IT Operations at the Forefront: Initial use cases are frequently centered on IT operations. Automated data classification, proactive storage optimization, and compliance reporting are already freeing up vital time and resources for IT personnel, while predictive maintenance and real-time cybersecurity responses significantly reduce downtime and risk. Agentic AI empowers proactive problem resolution, streamlining resource allocation, and bolstering resilience in hybrid environments.
Navigating the Southeast Asian Landscape: For Southeast Asian businesses, the initial focus should be on data quality and infrastructure readiness, as these foundational elements are critical for realizing the benefits of agentic AI. Strategically starting with IT operations can offer measurable success before expanding to other business domains.
Reshaping Workflows: Beyond IT: The transformative potential of agentic AI extends far beyond IT. Future applications in supply chain management and customer service are becoming increasingly conceivable. Imagine agentic AI anticipating capacity needs, balancing workloads, and performing predictive maintenance in supply chains, or providing superior personalized customer service experiences. Even cybersecurity is transformed, with agentic AI enabling rapid detection, isolation, and backup triggering, minimizing potential damages.
Skills & Leadership in the Age of Agentic AI: The rise of agentic AI necessitates a shift in human skills. Leaders need to effectively oversee and orchestrate autonomous systems, maintaining a balance between autonomy and accountability. This necessitates strategic investments in reskilling initiatives, which are already under way, demonstrating a crucial response to the potential workforce displacement.
The Future is Now: The economic and business model implications are profound. IDC projects significant contributions to ASEAN GDPs due to AI. The impact of agentic AI will likely be quicker and more substantial than currently anticipated across various sectors, not just confined to IT operations, but reshaping business structures, risk management, and value creation mechanisms. Successful adoption hinges on balancing autonomy with robust oversight, continuous reskilling, and a proactive approach, with data management and infrastructure forming the crucial foundational bricks.
Key takeaway: Agentic AI has massive potential in the enterprise, but successful deployment requires a proactive and holistic approach, emphasizing human oversight and strategic planning.