Unleashing agentic AI’s true potential: Strategic approaches for a silicon-based workforce

The agentic reality gap: Bridging promise and practice

Enterprises are rapidly exploring agentic AI, with Gartner predicting that 15% of daily work decisions will be made autonomously by agentic AI by 2028, a significant increase from virtually none in 2024. Despite this momentum, many organizations face substantial hurdles in translating agentic pilots into production-ready solutions. Deloitte’s 2025 Emerging Technology Trends study indicates that only 14% of organizations have deployable solutions, and just 11% are actively using these systems in production.

Several fundamental infrastructure challenges contribute to this gap. Traditional enterprise systems-often not designed for real-time, autonomous agent interactions-create significant bottlenecks, with Gartner forecasting that more than 40% of agentic AI projects will fail by 2027 due to legacy system incompatibility. In addition, data architectures built around batch-based Extract, Transform, Load (ETL) processes introduce friction for agent deployment. Many organizations also attempt to automate existing, human-centric processes instead of reimagining workflows for an agent-native environment, resulting in inefficient implementations and, in some cases, “workslop”—where poorly designed agentic applications actually increase operational burden.

Redesigning operations for a human–digital future

Leading organizations are addressing these challenges by adopting a systematic approach to agentic transformation, moving beyond simply layering agents onto legacy workflows. This requires stepping back to examine end-to-end processes and leveraging agents’ distinctive strengths, such as continuous, high-volume task execution without human constraints like breaks or working hours.

“Now is an ideal time to conduct value stream mapping to understand how workflows should work versus the way they do work. Don’t simply pave the cow path. Instead, take advantage of this AI evolution to reimagine how agents can best collaborate, support, and optimize operations for the business.”

This strategic redesign is already taking shape across industries. At HPE, an AI agent called Alfred supports internal operational performance reviews. Toyota uses agentic tools to gain real-time visibility into vehicle arrivals at dealerships and is planning to empower agents to resolve supply chain issues—bypassing the need for human interaction with complex mainframe systems. These examples signal a shift away from traditional application modernization toward enabling agents to bridge legacy system gaps directly.

Perhaps the most profound change is the recognition of agents as a new category of labor—a “silicon-based workforce.” Organizations are increasingly integrating these digital workers with their human, or carbon-based, workforce, enabling people to focus on higher-value activities such as governance, compliance, and growth strategy. Mapfre, for example, deploys AI agents to handle routine administrative tasks in claims management, while keeping humans in the loop for sensitive customer interactions. Moderna has gone a step further by merging its technology and HR functions under a Chief People and Digital Technology Officer to better integrate talent and technology.

“If we think of agents as digital skills, their real value emerges when they operate collectively. Most composite processes don’t exist solely within the enterprise. Trustworthy, secure interworking between agents is critical.”

Effective implementations also rely on specialized agents orchestrated at scale, automating entire workflows through emerging standards such as the Model Context Protocol (MCP), Agent-to-Agent Protocol (A2A), and Agent Communication Protocol (ACP). Together, these approaches enable a “microservices model for AI,” reducing complexity while supporting scalable orchestration and platform flexibility. At the same time, organizations are introducing FinOps frameworks for AI, essential for monitoring and controlling agent-driven costs, particularly in token-based pricing environments.

The shift toward agentic AI represents more than a technological evolution; it is a fundamental organizational transformation that will reshape how enterprises operate, compete, and create value. Organizations that master agent-native process design, multi-agent orchestration, and silicon workforce management will be best positioned to thrive in an increasingly automated economy. Ultimately, success will depend on creating new forms of human–AI collaboration that capitalize on the complementary strengths of both human and silicon-based workers.

To explore how organizations can strategically embrace agentic AI and prepare for the future of work, read the original article on the Deloitte Insights website.