Executive summary
AI creates capacity. Leaders must own the changes to work, authority and accountability that turn it into an accepted outcome. BCG’s effort model provides the context; Adaptation and Control make the operating implications concrete.
Intelligence is increasingly available on demand. Turning that capacity into a better operating result still requires choices about how an enterprise works.
BCG’s September 2026 report, The Formula for Agentic AI Value, sharpens that leadership challenge. In its research, 42% of companies expect to grant agents autonomy by 2030, while only 5% have the full set of critical controls in place today. Those figures describe reported expectations and readiness, not an inevitable destination for every organization. [1]
The report also reiterates BCG’s 10–20–70 model for transformation effort. The largest share concerns the people, organization and processes around the technology. That is where leadership attention needs to become operating commitment.
A faster task can leave the outcome unchanged. An invoice may be read instantly and still wait for approval. A recommendation may arrive sooner while the receiving team lacks the information, authority or capacity to act on it.
As systems gain permission to act, those dependencies become more consequential. Leaders need to know which decisions can be delegated, which conditions must hold and who intervenes when the operation moves outside its agreed limits.
The question is whether the enterprise is changing enough around the capability for its intended value to become real.
RePerspective Labs approaches this work through Adaptation and Control, grounded in technology and data.
Adaptation means redesigning the work around the outcome: connecting knowledge, removing unnecessary handoffs, changing responsibilities and helping people develop the judgment the new operation requires. It includes deciding how released capacity will be used. Time saved becomes valuable when the organization puts it to a worthwhile purpose.
Control makes delegated authority enforceable. Identity, permissions, action limits, monitoring and revocation need to work together. People must be able to intervene, recover the operation and answer for its consequences. Human accountability needs an operational design.
The enterprise AI value equation
BCG · Transformation effort
RPL · The operating interpretation
Redesign work and decisions. Connect knowledge. Develop people.
Enforce authority and limits. Monitor action. Enable intervention.
These responsibilities cross organizational and technical boundaries. A policy becomes useful only when the workflow, permissions and people can apply it. A redesigned process depends on reliable records and systems. The 70% deserves primary attention without treating the technology and data underneath it as someone else’s problem.
An operating experience
In a supplier-payment initiative, I defined the vision, architecture and design; my team delivered the solution. Extracting invoice details addressed part of the work. Eligibility also depended on work records, purchasing limits and the responsible manager’s recorded approval. When approval was missing, processing waited. This was a document-processing and validation workflow. I recall eligible invoices progressing more quickly, but do not present a measured performance claim. The Work Behind Faster Supplier Payments explains the design and its limits.
Fund the operating changes as part of the AI initiative. Give them named responsibility, delivery time and acceptance criteria alongside the technical work.
The Outcome owner answers for the end-to-end result, its acceptance criteria and funding. The Operating owner runs it day to day. Together with technology and control specialists, they need to establish what changes before wider use: the work, the authority, the information, the people’s responsibilities and the evidence of improvement.
This does not require redesigning the entire enterprise before making progress. Start with a useful boundary, understand the dependencies around it and demonstrate an accepted outcome. Then use the evidence to decide whether to Expand, Continue, Narrow, Redesign or Stop.
For your next AI investment, who owns the operating changes required to turn its new capacity into an accepted business outcome?
Explore the operating perspective behind RePerspective Labs.
Short, practical perspectives on AI-era operations, governance, and operating-model transformation.
Subscribe
RePerspective Labs