TechBay.News

By Michael Phillips | TechBayNews

Agentic AI — autonomous, goal-oriented systems that can plan, decide, and execute multi-step tasks with limited human input — is quickly emerging as the next major frontier in artificial intelligence. But new data suggests that while experimentation is widespread, meaningful enterprise-scale adoption remains rare.

That’s the central takeaway from a recent “Week in Charts” feature published by McKinsey & Company, which distills findings from its 2025 Global Survey on the State of AI into a single, revealing visualization. The chart shows where AI agents are actually being deployed at scale — and where most organizations are still struggling to move beyond pilots.

Broad Adoption, Shallow Impact

According to McKinsey’s survey of 1,993 respondents across 105 countries (conducted June–July 2025), AI adoption has reached near-ubiquity:

• 88% of organizations now use AI in at least one business function
• ~79% report using generative AI tools
• Over two-thirds deploy AI across multiple functions

Yet despite this momentum, nearly two-thirds of organizations have not begun scaling AI enterprise-wide. Most remain stuck in experimentation or limited deployments, with only incremental gains to show for it.

Agentic AI highlights this gap especially clearly.

Where Agentic AI Is Scaling Today

McKinsey’s bubble chart maps scaled adoption of AI agents by industry and function. The results show a clear pattern: scaling happens first where digital maturity is already high.

Technology leads the way

• Software engineering: ~24% report scaled agent use
• IT operations: ~22%
• Product and service development: ~18%

Insurance stands out in marketing and sales, while healthcare shows notable progress in knowledge management and IT, where structured data and repetitive decision flows are well-suited to autonomous systems.

Across all sectors, however, scaled adoption rarely exceeds the mid-20% range — underscoring how early the market still is.

High Performers Look Very Different

McKinsey’s full report, The State of AI in 2025: Agents, Innovation, and Transformation, produced by QuantumBlack, identifies a small group of “high performers” — roughly 6% of respondents — who attribute at least 5% of EBIT to AI.

These organizations are:

• 3x more likely to scale AI agents across functions
• Focused on growth and innovation, not just efficiency
• Redesigning entire workflows, not merely automating tasks
• Backed by senior leadership, often with CEO-level AI governance

In other words, success correlates less with which tools are chosen — and more with how deeply operations are reimagined.

Workflow Redesign Is the Real Differentiator

A recurring theme in McKinsey’s research is that agentic AI fails when it’s layered onto legacy workflows built for humans alone. Agents excel at planning, orchestration, and continuous execution — but only when processes are rebuilt around those strengths.

High-impact organizations rethink end-to-end flows such as:

• Lead-to-order
• Customer service resolution
• Hire-to-retire
• Compliance and reporting

In these redesigned models, agents handle diagnosis, coordination, and execution, while humans move “above the loop” for judgment, oversight, and empathy.

McKinsey vs. Gartner: Same Direction, Different Lens

McKinsey’s findings align closely with the 2025 AI outlook from Gartner, though the two approach the topic from different angles.

• McKinsey focuses on what’s happening now: broad adoption, limited scaling, uneven value capture.
• Gartner’s Hype Cycle places agentic AI near the Peak of Inflated Expectations, warning that over 40% of agentic AI projects could be canceled by 2027 due to unclear ROI, costs, or governance failures.

Both agree on one point: the shift from generative AI tools to autonomous agents is inevitable — but success will depend on discipline, redesign, and governance, not hype.

A Pivotal Moment for Enterprises

The takeaway for business leaders is increasingly clear. Agentic AI is no longer theoretical, but neither is it plug-and-play. The companies pulling ahead are those willing to rewire how work gets done — treating AI agents not as assistants, but as digital workers embedded into redesigned systems.

As McKinsey puts it, the question has shifted from “Which tasks should we automate?” to “Which workflows should we reimagine?”

For most organizations, the race to scale agentic AI is just beginning — and the winners will be decided less by experimentation, and more by execution.