A new Deloitte survey of 501 senior business leaders at US companies finds that most organizations are still struggling to move agentic AI from small trials into broad production use. The research shows that 42% of companies are testing small numbers of AI agents, while 43% are expanding deployments across functions. Only 15% have achieved scaled, orchestrated multi-agent deployments across customer service, IT, and engineering. Nearly two-thirds of leaders say they are reevaluating their business models because of advances in agentic AI, though only half have a clear view of their future operating model.
The biggest barriers to scaling agentic AI are not technical ambition but foundational gaps. According to the survey, 72% of leaders cite a lack of a unified, accessible data foundation as a key obstacle, while 70% point to an inability to trust and govern agents. Another 67% flag the cost and complexity of integration. The report also notes that technological change is outpacing organizations’ ability to adapt, framing the shift as a relational transformation rather than just a technology upgrade.
Workforce readiness is a particular weak spot. Only one in five businesses said their workforce is prepared for agentic AI, with a lack of employee reskilling and upskilling cited as a major hurdle. Fewer than a third of leaders said their organizations are ready in terms of vision and strategy, technology infrastructure, or data foundation, and only 26% said their risk, security, and governance frameworks are ready. Half of the leaders admitted their organizations are not investing enough in workforce transformation efforts.
Process redesign is also lagging. Just 16% of business leaders said their current processes are prepared for agentic adoption. The research warns against layering AI onto legacy workflows without rethinking them, though layering can serve as a bridge. Leaders expect significant change ahead: by 2030, 74% predict nearly half of business processes will be redesigned or rebuilt around AI agents, and 61% believe most processes will be powered by largely autonomous agents. However, only 31% expect such redesigns to happen by 2028.
The survey also highlights cost pressures and job disruption. Leaders anticipate new roles where humans and AI collaborate, but 43% expect major job disruption as routine tasks become automated. Training employees and managing token costs will add budget pressure, with 71% of organizations currently working on baseline AI agent literacy and 65% on upskilling and reskilling efforts.
According to Deloitte, scaling agentic AI requires an integrated roadmap, using layering as a temporary step, investing in workforce readiness, and defining a clear human-and-agent operating model. The report concludes that becoming an agentic enterprise means redesigning processes, reskilling labor, redeploying employees away from repetitive tasks, restructuring financial and organizational models, recalibrating metrics around agent usage, and rethinking leadership’s autonomous-focused strategy.
More AI news from TechManNews.
