Agentic AI is emerging as a transformative force in artificial intelligence, poised to revolutionize enterprise operations. Industry forecasts suggest that by 2028, agentic AI could generate up to US$450 billion in economic value. Despite this promising outlook, widespread adoption remains limited, with only 2% of organizations having fully integrated agentic AI solutions. Moreover, trust in autonomous AI agents is beginning to wane, highlighting a critical challenge for businesses.
Bridging the Gap Between Potential and Implementation
Capgemini’s recent study, conducted in April 2025 with 1,500 executives from large companies across 14 countries including Singapore, delves into this paradox of high potential but slow uptake. The research underscores that trust and governance are pivotal to unlocking agentic AI’s value. Nearly 75% of respondents emphasized that human involvement in AI-driven workflows delivers greater benefits than drawbacks, while 90% viewed oversight as either beneficial or neutral in cost.
The consensus is clear: agentic AI achieves optimal results when it complements human expertise rather than operating independently.
Current Adoption Landscape: Early Pilots and Gradual Progress
Approximately 25% of organizations have initiated pilot projects involving agentic AI, yet only 14% have advanced to full-scale implementation. Most companies remain in the planning phase, revealing a widening divide between ambition and operational readiness. This gap is a significant barrier to realizing the technology’s economic advantages.
Agentic AI is no longer a futuristic concept; practical applications are surfacing. For instance, consider a virtual personal shopping assistant capable of understanding detailed customer requests, generating tailored product descriptions, answering inquiries, and adding items to a cart via voice or text commands. Although these assistants typically halt short of completing payments due to security concerns, they effectively replicate many functions traditionally performed by human sales associates.
This evolution prompts a reevaluation of conventional e-commerce platforms. If AI can streamline product discovery, comparison, and purchase preparation, will consumers still need to navigate complex websites? For users overwhelmed by cluttered interfaces, AI-powered interactions could offer a more intuitive and accessible shopping experience.
Clarifying Agentic AI: Beyond Generative Models
To demystify agentic AI, Jason Hardy, Chief Technology Officer for Artificial Intelligence at Hitachi Vantara, explains its core attributes. “Agentic AI refers to software systems capable of autonomous decision-making, action-taking, and strategy refinement,” Hardy states. “Imagine a collaborative team of specialists that learns from experience, coordinates complex tasks, and operates in real time. While generative AI focuses on content creation in response to prompts, agentic AI actively pursues goals and adapts within dynamic environments.”
This distinction-between generating outputs and driving tangible outcomes-is fundamental to understanding agentic AI’s role in enterprise IT.
Drivers Behind Accelerated Adoption
Hardy identifies the increasing scale and complexity of enterprise environments as key catalysts for agentic AI adoption. “Organizations face mounting challenges related to complexity, risk, and operational scale. Agentic AI is gaining traction because it transcends mere analysis-it dynamically optimizes storage and capacity, automates compliance and governance, predicts failures before they occur, and responds to security threats instantaneously. This shift from insight generation to autonomous action is propelling faster adoption,” he explains.
Capgemini’s findings align with this perspective, revealing that although confidence in agentic AI varies, early implementations demonstrate significant value, particularly when automating routine yet critical IT functions.
Unlocking Value: Practical Applications in IT Operations
Hardy highlights IT operations as the most mature and impactful use case for agentic AI. “Automated data classification, proactive storage management, and compliance reporting save IT teams substantial time daily. Predictive maintenance and real-time cybersecurity interventions minimize downtime and mitigate risks,” he notes.
Beyond efficiency gains, these capabilities enable systems to identify issues proactively, allocate resources more effectively, and contain security breaches swiftly. “Early adopters are leveraging agentic AI to address incidents before escalation, enhancing system reliability and performance across hybrid cloud environments,” Hardy adds.
Given these measurable benefits, IT operations represent a pragmatic entry point for enterprises seeking to harness agentic AI’s potential.
Foundations for Southeast Asia’s Agentic AI Journey
For organizations in Southeast Asia, Hardy emphasizes the critical importance of robust data management. “Agentic AI’s effectiveness hinges on well-classified, secure, and governed enterprise data,” he explains.
Equally vital is the underlying infrastructure capable of supporting multi-agent coordination, persistent memory, and dynamic resource allocation. Without these technological foundations, agentic AI deployments will remain limited in scope and impact.
Many enterprises may find it strategic to begin with IT operations, where agentic AI can preempt system outages and optimize performance, before expanding into broader business areas.
Transforming Core Business Processes
Hardy envisions agentic AI reshaping workflows across IT, supply chain management, and customer service domains. “In IT, agentic AI can forecast capacity requirements, rebalance workloads, and reassign resources in real time. It also automates predictive maintenance, preventing hardware failures before they occur,” he explains.
Cybersecurity stands out as another promising field. “Agentic AI can detect irregularities, isolate compromised systems, and initiate immutable backups within seconds, drastically reducing response times and limiting damage,” Hardy observes.
These capabilities are moving beyond experimental stages, with early deployments already enhancing resilience and reliability in complex hybrid environments.
Human Capital and Leadership in the Age of Agentic AI
Successful integration of agentic AI demands new skill sets and leadership approaches. “The human role will evolve from task execution to oversight and orchestration,” Hardy predicts. Leaders must establish clear boundaries and continuously monitor autonomous systems to ensure ethical compliance and alignment with organizational goals.
For managers, this shift means focusing less on routine administration and more on coaching, innovation, and strategic planning. Human resources teams will need to develop governance competencies such as audit readiness and design frameworks for effective agentic AI integration.
The workforce impact will be uneven. The World Economic Forum forecasts that AI could create 11 million jobs in Southeast Asia by 2030 while displacing 9 million. Women and Generation Z workers are particularly vulnerable, with over 70% of women and up to 76% of younger employees in roles susceptible to AI-driven disruption.
This underscores the urgency of comprehensive reskilling initiatives. Significant investments are underway, including Microsoft’s $1.7 billion commitment in Indonesia and extensive training programs across Malaysia and the region. Hardy stresses that capacity building must be inclusive, swift, and strategically targeted.
Looking Ahead: The Future Impact of Agentic AI
Projecting three years forward, Hardy warns that many executives may underestimate the speed of transformation. “The initial wave of benefits is already evident in IT operations, where agentic AI automates data classification, storage optimization, predictive maintenance, and cybersecurity responses, freeing teams to concentrate on higher-value strategic activities,” he says.
However, the broader economic and business model implications could be even more profound. Projections indicate that AI and generative AI could contribute approximately US$120 billion to the GDP of the ASEAN-6 countries by 2027. Hardy believes the pace and scale of impact will surpass current expectations.
In Indonesia alone, over 57% of job roles are anticipated to be augmented or disrupted by AI, signaling that transformation will extend well beyond IT, influencing organizational structures, risk management, and value creation.
Striking the Right Balance: Autonomy Meets Accountability
Both Capgemini’s research and Hardy’s insights converge on a vital theme: while agentic AI offers immense promise, its successful application depends on harmonizing autonomous capabilities with human trust and oversight.
When implemented thoughtfully, agentic AI can reduce operational costs, enhance system reliability, and unlock new revenue opportunities. Yet, without robust governance frameworks, targeted reskilling, and infrastructure readiness, adoption risks stagnation.
For Southeast Asian enterprises, the critical question is not if agentic AI will become mainstream, but how swiftly it will be embraced-and whether organizations can maintain accountability as AI systems assume greater responsibility in business decision-making.
