KPMG: Transforming the Enterprise 2026

13 Oct

Source: KPMG, Transforming the Enterprise: Reframing how enterprises execute, scale and sustain change, June 2026

Key takeaways

Transformation activity is expanding faster than enterprise performance. Nearly every organisation is transforming and the average organisation is managing 3.5 transformations concurrently, yet only 14% consider themselves top performers relative to industry peers.
The challenge has shifted from initiating transformation to orchestrating it. KPMG argues that the limiting factor is increasingly the ability to align priorities, integrate execution and maintain control across interconnected transformation activity.
Early AI productivity gains are not the same as enterprise transformation. Many organisations are generating local productivity improvements while applying AI to existing workflows, operating models and decision structures. Sustained value requires those gains to reshape how work is performed, decisions are made and capabilities connect across the enterprise.
Most organisations know which capabilities matter but have not built them at scale. Foundational technology and data and customer engagement are each considered critical by 61% of respondents, yet only 13% regard themselves as leaders in either capability.
AI deployment is moving faster than operating model change. KPMG argues that AI value often remains concentrated in personal productivity because organisations have not integrated AI deeply enough into enterprise workflows, decision making and operating models.
Technology foundations need to evolve into an intelligence layer. AI increasingly relies on shared data, enterprise context, interoperable systems and a common understanding of how customers, products, processes, risks and decisions connect.
Trust becomes operational rather than a downstream control. Governance, security and accountability increasingly need to operate inside decision making and execution as AI systems become more autonomous.
Only 24% of organisations have proactively integrated AI risk management into strategy and the technology lifecycle. Despite 60% viewing trust as a strategic differentiator or competitive advantage, only 28% measure operational or revenue outcomes linked to trusted AI.
The operating model has become a strategic constraint. Only 12% of organisations can move a new initiative from concept to execution in under three months, while 51% require six months or longer.
The next performance frontier is human AI coordination. 34% of leaders expect the largest operational performance gains over the next three to five years to come from combining AI agents with human decision making, ahead of AI agents alone at 25%.
Agentic AI is spreading faster than work is being redesigned. Deployment now extends beyond technology into marketing and sales, operations, risk, finance and HR, but roles, decision rights and accountability often remain largely unchanged.
Only 19% report a highly AI proficient workforce. KPMG argues that closing the gap requires more than training because organisations must also rethink how work is organised, how decisions are made and how accountability is structured.
Human AI integration is promoted more often than it is operationalised. 75% say leadership promotes integration of AI and human intelligence, but only 40% strongly agree. Similar gaps appear around human oversight, continuous learning and cross functional AI integration.
Orchestration is distinct from coordination. Coordination relies on forums, committees and alignment. Orchestration establishes decision rights, intervention thresholds, stopping rules and mechanisms for dynamically reallocating capital and capability.
Transformation leaders do not necessarily transform more. They prioritise adaptability over efficiency, enterprise alignment over transformation volume, alignment before scaling AI, trust as an enabler of speed and external ecosystems as part of execution.
The emerging model is continuous enterprise orchestration. Strategy, technology, operations, workforce and governance increasingly need to operate as one adaptive system rather than separate transformation programmes.

Executive summary

KPMG’s Transforming the Enterprise 2026 examines why the large volume of AI, digital and operating model transformation now underway is not consistently translating into stronger enterprise performance.

The research surveyed 1,750 senior transformation leaders across 20 countries and territories. The central finding is that organisations do not appear to lack ambition, technology or transformation activity. The problem is execution across an increasingly complex enterprise.

Only 1% of organisations report that they are not currently transforming. Ninety four percent are undertaking at least two transformations simultaneously, 72% are running three or more and the average organisation is managing 3.5 concurrent transformations. Yet only 14% consider themselves top performers relative to their industry peers.

KPMG argues that the fundamental issue is that organisations continue to treat transformations as separate initiatives while the underlying enterprise is becoming increasingly interconnected. AI, technology modernisation, workforce change, operating model redesign, risk and business model transformation now influence one another. The more initiatives organisations launch, the greater the coordination problem becomes.

AI intensifies both the opportunity and the problem. Many organisations are already generating meaningful productivity gains, but those gains frequently remain local. AI is often layered over existing technology architectures, workflows and decision structures rather than used to redesign the underlying enterprise.

KPMG therefore distinguishes transformation activity from enterprise orchestration. Coordination relies on governance meetings, alignment forums and consensus mechanisms. Orchestration goes further by establishing decision rights, intervention thresholds and stopping rules while dynamically reallocating resources towards the activity creating the greatest enterprise value.

This leads to five priorities. Organisations need to rebuild technology foundations around an integrated intelligence layer, establish trust as an operating capability, redesign operating models around end to end value streams, redesign work for human AI collaboration and rethink transformation itself as continuous enterprise orchestration.

The technology challenge starts with fragmented foundations. Most enterprises operate across combinations of SaaS platforms, proprietary systems, legacy architecture and increasingly diverse AI capabilities. AI can generate value inside individual functions, but that value does not automatically compound across the enterprise.

KPMG therefore argues for an intelligence layer connecting data, workflows, decisions and business context. A shared enterprise ontology becomes increasingly important because AI needs more than raw data. It needs a consistent representation of how customers, products, processes, risks, policies and decisions relate to one another.

Trust also changes role. Governance models built around periodic review struggle when intelligent systems adapt and execute continuously. Trust therefore needs to move into the operating model itself so risk, governance, security and accountability shape what intelligent systems can do in real time.

The same redesign is required for work. Current workforce models were largely built for deterministic processes where tasks, decision rules and expected outcomes are relatively predictable. AI is probabilistic. It learns from context and produces recommendations based on inference. KPMG argues that organisations cannot simply insert probabilistic technology into a deterministic workforce model and expect enterprise transformation.

The result is a shift towards what KPMG calls the human AI enterprise. Work becomes organised more around end to end value streams, AI increasingly supports analysis, coordination and execution, and people concentrate more heavily on judgement, oversight and innovation.

Management changes with it. Instead of coordinating only human contributors, managers increasingly orchestrate a broader workforce of employees, AI agents, automation, partners and platforms.

The report ultimately argues that the next transformation advantage will not come from how much transformation an organisation undertakes. It will come from how effectively the organisation directs, connects and continuously adapts it.

Key stats and quotable claims

The study surveyed 1,750 senior transformation leaders across 20 countries and territories.
Only 1% of organisations report that they are not currently undertaking transformation.
94% are undertaking two or more transformations simultaneously.
72% are undertaking three or more transformations.
54% are undertaking four or more.
36% are undertaking five or more.
Organisations are managing an average of 3.5 concurrent transformations.
Only 14% consider themselves top performers relative to industry peers.
Only 26% strongly agree that AI has positively contributed to their growth objectives.
Foundational technology and data capabilities are considered critical by 61%, but only 13% say they lead in this capability.
Customer engagement capabilities are considered critical by 61%, but only 13% say they lead.
Enterprise wide operating capabilities are considered critical by 58%, but only 12% report leading maturity.
Modern internal support services are considered critical by 55%, while only 11% report leadership.
Predictive operational capabilities are considered critical by 55%, with only 11% leading.
Risk transformation is associated with an average self reported performance improvement of 14%.
Digital and technology transformation is associated with 9%.
Business model transformation is associated with 8%.
Regulatory transformation is associated with 8%.
Functional transformation is associated with 5%.
70% of organisations report active digital transformation.
51% report operating model transformation.
48% report enterprise wide change across systems, processes and people.
AI is fully embedded in customer analytics and personalisation in only 27% of organisations.
AI is fully embedded in operational efficiency activities in 23%.
AI is fully embedded in strategic decision support in 26%.
Only 19% report AI fully embedded in developing new revenue streams or business models.
Among organisations reporting AI productivity gains, 48% describe the improvement as incremental and 40% as substantial.
39% measure AI value through increased productivity.
36% measure time saved.
33% measure operational cost reduction.
Only 22% measure revenue from AI driven business models.
43% operate a hybrid AI ecosystem combining third party platforms with internally developed capability.
Only 24% have proactive and integrated AI risk management embedded into strategy and the technology lifecycle.
45% respond to unexpected AI outcomes by handling them inside technical teams without wider communication.
Only 32% treat unexpected AI outcomes as learning opportunities with transparent review.
60% see trust as either a strategic differentiator or core competitive advantage.
Only 28% measure operational or revenue outcomes linked to trusted AI.
Only 12% can launch new initiatives from concept to execution in under three months.
51% require six months or longer.
87% say operational agility creates a competitive advantage.
The leading obstacle to AI ROI is high implementation cost and unclear ROI at 29%, followed by technical and data infrastructure challenges at 23% and talent shortages at 22%.
34% expect the greatest operational gains over the next three to five years from combining AI agents and human thinking.
25% expect the largest gains from AI agents alone.
Agentic AI is currently being deployed in 59% of technology and IT functions, 43% of marketing and sales, 42% of operations, 37% of risk and compliance, 33% of finance and 31% of HR.
Only 19% describe their workforce as highly proficient in using AI.
75% agree leadership promotes integration of AI and human intelligence, but only 40% strongly agree.
72% agree clear roles exist for human oversight, while only 35% strongly agree.
67% agree employees receive continuous AI training, while only 35% strongly agree.
Transformation leaders are nearly nine times more likely to be leading in enterprise capability maturity.
They are three times more likely to significantly outperform competitors.
They are more than twice as likely to treat trust as a source of competitive differentiation, at 70% versus 31%.
HR is moving towards total workforce management in 61% of organisations, yet only 11% say they have fully transitioned.
94% report at least partial AI adoption in finance.

Overall summary

KPMG’s central argument is that enterprise transformation has entered a different phase. Transformation itself is no longer unusual. Organisations are simultaneously modernising technology, adopting AI, redesigning operating models, responding to regulation, transforming functions and exploring new business models.

The challenge is that most enterprise structures were designed to coordinate a more stable environment. As transformation becomes continuous and increasingly interconnected, existing systems of governance, management and resource allocation begin to struggle.

KPMG therefore reframes transformation as an orchestration problem.

The difference matters. Traditional transformation management asks whether individual programmes are on track. Enterprise orchestration asks whether the programmes reinforce one another, whether decision rights are clear, whether resources remain aligned with the highest value opportunities and whether leadership can intervene as conditions change.

AI exposes the problem particularly clearly. Organisations are already generating productivity improvements, but much of the value remains concentrated in local use cases. Productivity can improve while the broader organisation remains unchanged.

KPMG argues that enterprise value begins to compound when AI becomes part of the operating model. Data and technology need to connect through an intelligence layer. Workflows need to cross functions. Business context needs to become legible to intelligent systems. Governance needs to operate in real time. Human and AI roles need to be deliberately designed.

The workforce problem is therefore not simply one of AI skills. Only 19% describe the workforce as highly proficient, but KPMG explicitly argues that training alone will not solve the mismatch. Organisations need to redesign work, decision rights and accountability around the different characteristics of human and machine intelligence.

That redesign shifts the role of management. Management moves from coordinating individuals towards orchestrating combinations of people, agents, automation and partners. Similarly, transformation leadership moves from launching initiatives towards deciding what should scale, what should stop and where resources should move.

The resulting enterprise is increasingly managed as an integrated system.

Deep dive

Enterprise transformation has entered a new phase

Enterprise transformation has entered a new phase. Intelligent economies are reshaping how organizations operate, demanding greater speed, adaptability and coordination than traditional models were designed to support. The challenge is no longer initiating transformation, but orchestrating it: aligning priorities, integrating execution and maintaining control as complexity accelerates across functions, technologies and operating models.

As intelligent systems extend across the enterprise, organizations are becoming more dynamic, interconnected and able to respond in real time. AI is no longer simply a standalone technology initiative. It is increasingly embedded in decision-making, execution and value creation. Advantage now depends on operating as an intelligent enterprise: one that can adapt continuously, grow sustainably, operate efficiently, modernize at pace and build trust at scale.

Most organizations, however, are not designed for this reality. They are increasing transformation activity without redesigning how the enterprise operates to absorb it. As a result, complexity is compounding faster than performance — slowing execution, constraining growth and eroding accountability.

KPMG’s argument is deliberately broader than technology transformation. Delivering sustained performance requires strengthening the enterprise as an integrated system. Leaders must rebuild technology and data foundations to enable AI at scale, while establishing trust, governance and resilience that accelerate confident adoption. At the same time, they must redesign how work flows end-to-end across value streams, embedding human-AI collaboration into how the enterprise operates and adapts.

In an intelligent economy, intelligent enterprises are better able to compete. They are designed as integrated, adaptive systems that align decision-making, execution and value delivery. Transformation cannot sit at the edges; it must be aligned and directed from the center.

Orchestration consequently becomes a defining source of competitive advantage. Organizations that can coordinate strategy, technology, operations and talent as an integrated system should be better positioned to sustain performance, accelerate growth and adapt to continuous change. Those that cannot are likely to struggle to keep pace with increasing complexity.

Early AI gains expose the next transformation challenge

KPMG’s executive summary begins with a clear diagnosis: enterprise transformation is not yet delivering on its promise. Not for lack of ambition, investment or technology, but because most organizations are scaling change without redesigning how the enterprise operates.

The defining challenge is no longer generating transformation activity, but exercising the control required to translate fragmented efforts into integrated execution across the enterprise. Investment in AI, digital capabilities and new operating models continues to accelerate. Transformation is now widespread, spanning functions, systems and business units simultaneously, with most organizations managing multiple initiatives at once and many undertaking three or more transformations concurrently.

Despite significant activity and investment, relatively few organizations believe they are achieving transformation outcomes at scale. Only 14 percent of organizations consider themselves top performers relative to their peers, and only 26 percent strongly agree that AI has contributed positively to achieving their growth objectives.

This gap reflects a deeper breakdown in enterprise execution. Organizations are increasing transformation activity, but many still struggle to align priorities, coordinate decisions and sustain execution as change accelerates. As a result, transformation activity is expanding faster than enterprise performance.

At the same time, organizations are realizing meaningful productivity improvements from AI-driven initiatives. These gains demonstrate AI’s potential to improve operational performance, accelerate execution and augment decision-making. However, productivity gains alone do not constitute transformation.

Many organizations are applying AI to existing operating models, workflows and decision structures without fundamentally redesigning how the enterprise operates. What differentiates leading organizations is not the speed of early gains alone, but whether those gains reshape how work is performed, how decisions are made and how capabilities connect across workflows and operating models to drive sustained enterprise performance.

The challenge is therefore no longer initiating transformation. It is orchestrating it.

The gap between activity and performance

Despite widespread transformation activity, enterprise performance gains remain concentrated among a relatively small group of organizations. Only 14 percent consider themselves top performers relative to their peers.

Most leaders understand what is required to compete, yet only a small minority believe their organizations execute these capabilities effectively at scale. The research shows that capabilities such as data, governance, workforce readiness and AI enablement are viewed as critical to competitive advantage. The limiting factor is no longer ambition, but the ability to build, integrate and continuously evolve these capabilities as part of how the organization operates.

The difference lies in how capabilities are developed and scaled. In leading organizations, capability building is embedded into day-to-day operations rather than treated as a series of discrete transformation initiatives. This allows strategic priorities to translate into coordinated execution as business conditions evolve.

What differentiates leading organizations is not the scale of transformation activity, but how effectively transformation becomes part of how the enterprise operates.

The capability gaps themselves are large. Foundational technology and data capabilities are considered critical to future advantage by 61% but only 13% say they lead. Customer engagement shows the same 61% versus 13% split. Enterprise wide operating capability is considered critical by 58% while just 12% report leadership, and predictive operational capability shows a 55% versus 11% gap.

The issue is therefore not awareness. Enterprises broadly know what needs to be built. The challenge is converting strategic recognition into integrated capability at scale.

Autonomous systems are putting governance under strain

As systems become more automated, existing approaches to governance are coming under increasing strain. Static controls and periodic oversight struggle to keep pace with environments that evolve continuously and adapt in real time.

Most governance models were designed for environments where decisions were made by people, processes changed incrementally and accountability could be traced through relatively stable structures. Those assumptions are becoming less reliable as intelligent systems take on a greater role in decision-making and execution.

Visibility diminishes, accountability becomes more complex and risks can compound more quickly. Governance should evolve to become more embedded, adaptive and responsive to how decisions are actually made across the enterprise.

This is where KPMG introduces orchestration as the defining capability of a continuously evolving enterprise. From its perspective, the organizations pulling ahead are not simply those deploying AI the fastest; they are those building the capabilities that make transformation work.

At the center of this shift is orchestration: the enterprise capability to align priorities, integrate execution and maintain control across interconnected activities.

Coordination is not orchestration

KPMG makes an explicit distinction between coordination and orchestration.

Coordination relies on alignment mechanisms: committees, governance forums and consensus-building across functions. While necessary, these approaches are typically insufficient to deliver transformation at scale.

Orchestration operates differently. It establishes decision rights, defines intervention thresholds, enforces stopping rules and dynamically reallocates capital and capability to where they can create the greatest value.

Without this shift, transformation efforts often remain fragmented, competing for resources rather than reinforcing one another. Orchestration helps ensure that initiatives are actively directed across the enterprise so that value can compound rather than dissipate across disconnected efforts.

The report therefore defines orchestration as the enterprise capability to align priorities, integrate execution and maintain control across interconnected activities.

This is more than better programme management. It changes who can make decisions, when leadership intervenes, when an initiative should stop and how capital and capability move as conditions change.

Three conditions for scaling transformation

KPMG argues that scaling transformation requires organizations to fundamentally rethink how the enterprise is built, operated and governed.

The first requirement is to rebuild the foundations. Organisations need to strengthen trust and governance across data, platforms, security and risk so the enterprise can scale with confidence as complexity and uncertainty increase.

The second is to redesign work. Organisations need to reshape workforce design and ways of working to embed a human-AI operating model, integrating roles, workflows and decision-making to help improve productivity and effectiveness.

The third is to rethink the enterprise. Leaders need to evolve the execution model to align capabilities, functions, systems and decisions so the organization can direct activity, adapt quickly and deliver consistently at scale.

These three areas form the structure for much of the report.

Five strategic priorities

KPMG then translates the argument into five priorities.

The first is to rebuild technology foundations for an intelligence layer. Leaders should prioritize rebuilding core technology foundations to support AI as a permanent enterprise capability rather than a series of pilots. This requires investing in secure, resilient and interoperable platforms where data, systems and AI models are designed to work together from the outset.

Rather than layering AI onto disconnected systems, organizations should establish a shared, integrated core that enables intelligence to flow across the enterprise, supporting faster innovation, clearer accountability and more scalable execution.

The second is to establish trust foundations that enable speed and value. Trust should be embedded directly into how the enterprise operates rather than treated as a control applied after deployment. Governance, risk management, safety and security should be integrated into decision-making and execution so AI can be relied upon at scale.

When trust is treated as a performance enabler — actively managed, reinforced through behavior and embedded into operations — organizations can move faster with confidence, reduce friction and scale AI responsibly.

The third is to redesign the operating model around end-to-end value streams. Performance increasingly depends on how work flows across the enterprise rather than within individual functions. Organizations should redesign operating models around end-to-end value streams, aligning strategy, customer needs, operations and technology into a more unified execution model.

By reducing handoffs and clarifying accountability, leaders can accelerate decisions and help improve outcomes across the full flow of value.

The fourth is to reimagine work for human-AI collaboration. One of the next productivity frontiers lies in how people and AI operate together by design. Organizations should redesign roles, workflows and decision-making to integrate human judgment and AI capabilities into daily work.

This requires embedding continuous learning into the flow of work, with clear accountability and incentives aligned to new ways of operating, helping to ensure workforce capability evolves alongside technology rather than constraining it.

The fifth is to rethink enterprise transformation as agile enterprise orchestration. Transformation is increasingly difficult to manage as a collection of independent initiatives. The ability to continuously align priorities, coordinate execution and direct resources across the enterprise as conditions evolve is critical to success.

This typically requires clear decision rights, active management of trade-offs and real-time visibility into performance and dependencies.

Enterprise transformation in the age of AI

Transformation is becoming continuous and enterprise wide

Transformation is now nearly universal. Only one percent of organizations report that they are not currently undertaking transformation, while most are managing multiple initiatives simultaneously across AI, digital capabilities, operating model redesign and new business models.

As these initiatives expand across functions, systems and business units, they increasingly intersect and influence one another. What was once episodic is becoming more continuous, creating new connections across the enterprise.

These growing interdependencies can create new coordination challenges. Initiatives are no longer independent, increasing pressure to align decisions, manage trade-offs and coordinate execution.

Ninety four percent of organisations are undertaking at least two transformations, 72% are undertaking three or more, 54% are managing four or more and 36% are managing five or more. The average is 3.5 concurrent transformations.

The implication is that transformation management itself needs to change. A structure designed to manage one large programme at a time becomes increasingly ineffective when several interconnected transformations are simultaneously changing technology, workforce, governance and the operating model.

Misalignment becomes a primary constraint

As organizations manage concurrent transformation efforts, duplication, competing priorities and inconsistent execution often increase. The challenge is no longer just initiating transformation, but also aligning strategy, operating models, technology, workforce capability and governance into a coherent enterprise system.

Artificial intelligence is intensifying both opportunity and complexity. AI is accelerating the pace and scope of transformation. As it becomes embedded across workflows, decisions and customer interactions, organizations can operate faster and more intelligently, but also become more interconnected and dependent on coordinated execution.

Many organizations are generating localized gains without redesigning the workflows, operating models and governance structures required to scale AI consistently end-to-end.

Leaders are also operating in environments defined by uncertainty, conflicting signals and heightened expectations, yet clarity on where to focus and how to sequence transformation often remains limited.

As a result, transformation is frequently driven as much by urgency as by strategic intent, which can lead to activity that is not consistently aligned with enterprise outcomes.

Enterprise orchestration offers a path forward

Transformation can no longer be managed as a portfolio of independent initiatives. As organizations manage growing numbers of interconnected change efforts, the ability to align priorities, coordinate execution and maintain visibility across the enterprise becomes increasingly important.

Enterprise orchestration offers a way to connect strategy, execution and governance across these activities, helping organizations navigate complexity while sustaining performance.

KPMG’s Bayesian analysis reinforces the idea that no one enterprise capability independently explains transformation success.

The analysis shows that transformation success is not driven by any single capability, but by how capabilities connect and reinforce one another to drive enterprise performance.

The first finding is that there is no single path to successful AI transformation. Organizations can achieve strong outcomes through different combinations of capabilities, leveraging their unique strengths and following different pathways to success.

The second is that although pathways vary, successful transformations tend to share common characteristics. The strongest outcomes emerge when interconnected capabilities reinforce one another in ways that strengthen adoption, accelerate impact and enable AI to scale more effectively across the enterprise.

Three pathways towards AI performance

KPMG identifies three illustrative pathways.

Pathway one combines leadership commitment, culture and openness and use case strategy. Organizations with strong leadership commitment are more likely to create cultures that support experimentation, AI adoption and coordinated execution. These conditions help translate transformation ambition into clearer use case strategies and stronger success in embedding AI across the enterprise.

Pathway two combines talent depth, data readiness and training. Organizations with deep workforce capabilities are more likely to strengthen data readiness and invest in training that supports AI adoption and execution. Together, these conditions can create the operational foundation required to successfully embed AI at scale.

Pathway three combines strategic alignment, governance and cybersecurity. Organizations with strong strategic alignment are more likely to reinforce governance and cybersecurity capabilities across the enterprise, helping create the trust and stability required to successfully embed AI at scale.

The most effective pathway depends on an organization’s existing strengths and capabilities, sector dynamics, regulatory pressures and digital maturity. Choosing that pathway often means focusing investments on the capabilities that reinforce one another and create momentum for AI adoption, impact and scale.

Different strengths. Different pathways. Shared outcomes.

The enterprise transformation agenda is an integrated capability system

Transformation at scale depends on a set of interconnected capabilities, with orchestration at the center of this system to align priorities, integrate execution and sustain performance.

The first capability is AI technology foundations for integrated data and the intelligence layer. The strategic outcome is enterprise-wide integration of data, technology and business context, helping AI operate reliably and scale across use cases. The operating requirements include modular architecture, interoperable systems, data integration and governance and business ontologies supporting AI decision-making.

The second is trusted foundations for speed and value. The aim is transformation at scale with confidence, helping ensure innovation and AI adoption do not outpace governance and oversight. This requires integrated governance, responsible AI and data practices, cybersecurity, resilience and risk management aligned from the outset.

The third is a value stream based operating model, intended to create faster, more integrated execution across functions, systems and decisions. That requires end-to-end workflow redesign, simplification of operating models, cross functional integration and alignment of decision rights.

The fourth is human-AI collaboration, producing productivity and decision quality at scale through new ways of working that integrate human expertise and AI capabilities. Role and workflow redesign, workforce capability building, continuous learning and AI embedded into decision making become essential.

The fifth is agile enterprise orchestration, creating enterprise-wide direction, alignment and control across strategy, execution and performance. This requires enterprise-level control of priorities and decisions, cross-functional leadership alignment, dynamic resource allocation and real-time visibility into performance and trade-offs.

Rebuilding the foundations

Why AI can fail to scale

Most organizations are not yet designed to scale AI effectively. Rather than operating as integrated systems, they are layering AI capabilities onto fragmented technology and data foundations that were not built for enterprise-wide intelligence.

Leading organizations are taking a different approach. They are rebuilding their technology foundations to enable AI at scale — replacing fragmented systems and legacy applications with secure, resilient ecosystems designed for continuous change.

In these organizations, cloud platforms, modern data architectures and agentic AI operate as a connected ecosystem. Technology is designed for resilience, security and interoperability, enabling intelligence to flow across the enterprise and support faster innovation, stronger collaboration and more dynamic responses to changing market conditions.

AI adoption is accelerating across organizations. Investment is increasing, use cases are expanding and capabilities are being introduced across functions, workflows and decision-making.

Yet many organizations are still struggling to translate these gains into coordinated enterprise-wide impact. While AI is often creating value in specific functions and use cases, those gains are not consistently translating into enterprise-wide performance. As a result, value is often created locally rather than compounded across the enterprise.

Hybrid ecosystems increase complexity

Many organizations now operate across hybrid ecosystems, combining Software as a Service platforms, proprietary systems and AI capabilities without a consistent approach to managing how these components interact.

While this expands access to capability, it can also make it harder to maintain visibility, coordinate execution and ensure consistent enterprise performance.

As AI scales across these environments, the challenge becomes less about deploying capability and more about managing how it operates collectively. Solving this is essential to moving from isolated gains to sustained enterprise performance.

Despite widespread deployment, AI remains unevenly embedded within day-to-day operations and decision-making. More than half of organizations report broad AI use across activities such as customer analytics, operational efficiency and decision support, but only a minority report that these capabilities are fully embedded within operational workflows.

The difference is particularly visible across functions and sectors. Technology and supply chain report higher levels of operational embedding, while finance and HR continue to lag.

The intelligence layer becomes pivotal

As AI adoption accelerates, organizations are discovering that deploying new capabilities alone is often insufficient to scale AI effectively. An enterprise-wide intelligence layer helps connect decisions, workflows, data and execution across the business.

Without this unifying mechanism, capabilities often expand faster than they can be integrated with other areas of the business. As a result, value often remains localized rather than scaling consistently across the enterprise.

Many organizations remain caught between experimentation and enterprise transformation, with much of the value realized today concentrated in personal productivity gains such as automating tasks, accelerating insights and reducing manual effort.

These gains are meaningful, and many organizations report significant productivity improvements from AI-driven initiatives. The next phase of transformation is not simply generating more productivity, but embedding these capabilities into the operating model so that value can be coordinated, scaled and sustained.

As Adrian Clamp puts it in the report: “Most organizations have an AI value creation problem. Their AI deployment is moving faster than their pace of operating model change.”

Productivity remains the dominant measure of AI value

In evaluating AI’s value, most organizations continue to focus on operational metrics such as productivity at 39%, time savings at 36% and cost reduction at 33%. Far fewer measure strategic outcomes such as revenue growth, competitive advantage or new business models.

Operational gains are important indicators of progress, but they do not represent the full potential of enterprise value creation. Without integrated systems and coordinated execution, AI can improve efficiency without fundamentally strengthening competitive advantage.

The report therefore distinguishes between AI generating measurable local productivity and AI changing enterprise economics.

If the organisation measures primarily time saved, tasks automated or local cost reduction, it can miss whether those gains changed revenue, resilience, customer outcomes or competitive positioning.

The dominant architecture is becoming hybrid

As organizations scale AI, they are shifting away from static architectures toward more modular, adaptive technology environments designed for continuous change.

Forty-three percent of organizations report operating through hybrid ecosystems that combine third-party platforms with internally developed capabilities. While this model can provide greater flexibility and speed, it can also increase the importance of integration and governance more holistically.

Hybrid ecosystems often increase coordination complexity. As AI ecosystems expand, platforms, tools and models often operate with different assumptions about data, workflows and governance, increasing the potential for fragmentation across the enterprise.

Coordinating these environments requires greater integration across systems, functions and decision-making processes. These challenges can become more pronounced when adoption progresses unevenly across the organization.

Technology and executive leadership teams report higher levels of AI proficiency and integration while functions such as finance, risk and HR continue to lag, creating additional barriers to coordination and enterprise-wide scaling.

Existing systems are being extended rather than replaced

Most organizations are not replacing systems like enterprise resource planning, customer relationship management or service platforms. Instead, they are extending them with intelligence layers that interpret data, generate insight and support decision-making.

This can create new opportunities for innovation, but can also increase the importance of integration, orchestration and governance across systems and workflows.

In heavily regulated industries such as financial services and healthcare, modernization should also support trust, resilience and regulatory alignment from the outset.

As AI capabilities scale across increasingly interconnected environments, modernization is no longer just about simplifying systems. It is also about creating the foundations required to align data, decisions, workflows and governance across the enterprise.

This can require rebuilding data integration and pipelines, decision logic and orchestration and governance across operations and analytical layers. When these foundations are aligned, organizations can modernize more flexibly while enabling systems to continuously learn, adapt and scale.

Data alone is not enough

AI is only as effective as the data and business context that supports it. In many organizations, data remains siloed, inconsistently governed and difficult to access in usable forms.

However, data alone is not enough.

AI also relies on a shared understanding of how customers, products, processes, risks and decisions connect across the enterprise.

Without this shared context, AI can process information but cannot consistently apply it in ways that align with business priorities, constraints and operating realities.

This challenge becomes even more complex across large enterprises operating across multiple business units, regions and regulatory environments where standards, governance models and operating structures often differ significantly.

Shared enterprise context requires ontology

At the core of this challenge is the need for a shared enterprise ontology: a consistent model defining how key elements of the business relate to one another.

This can create a common operating language across systems, functions and intelligent agents, which can reduce fragmentation and improve coordination across the enterprise.

Traditional enterprise architecture focused primarily on systems integration. KPMG argues that intelligent enterprises require something broader: an architecture that connects data, workflows, decisions, policies and technologies into a coordinated operating environment.

As organizations move toward more adaptive and event-driven operating models, work should be coordinated dynamically rather than through static processes and manual intervention.

Organizations that establish this foundation are typically better positioned to scale AI consistently, coordinate execution more effectively and adapt more quickly as conditions change.

Trust: a foundation for scaling transformation

Trust is becoming operational

As AI becomes embedded into how work gets done, trust should shape how the enterprise operates in real time. Governance, security and accountability are no longer downstream considerations — they influence how decisions are made and executed across the business.

Organizations can struggle if they rely on governance models built around periodic oversight and downstream controls. As systems become more autonomous and interconnected, trust should be embedded directly into the flow of decision-making and execution across the enterprise.

Many organizations still treat governance, risk and security as functions applied after deployment rather than capabilities embedded into transformation itself.

As a result, the push for speed often outpaces the systems required to govern it consistently, which can create friction across execution, slow decision-making and reduce confidence in AI adoption at scale.

Organizations reporting stronger performance outcomes are more likely to embed governance and accountability directly into how decisions are made and work gets done.

In these organizations, trust appears to function as more than a safeguard, supporting execution, coordinated transformation and adaptability as complexity increases.

Most organisations have not embedded trust

Despite its importance, trust is not consistently embedded into how transformation is designed and executed.

Only 24 percent of organizations have proactively integrated risk management into strategy and the technology lifecycle, while most rely on partially integrated, reactive or decentralized approaches.

The sector pattern varies. Technology intensive and asset heavy industries such as energy, automotive and telecommunications report higher levels of proactive, integrated governance. Highly regulated sectors such as financial services also tend to embed risk and governance earlier in the lifecycle, often as a requirement rather than a choice.

But integration remains far from universal. In many organisations, the push for speed continues to outpace the systems required to govern it.

Leadership behaviour determines whether AI failures become learning

The research reveals meaningful differences in how organizations operationalize trust. Organizations that report stronger performance outcomes place greater value on proactive governance, trust management and the measurement of trust-related outcomes.

Rather than viewing trust primarily as a mechanism for managing risk, these organizations are more likely to embed it directly into decision-making, execution and transformation efforts.

When AI produces unexpected outcomes, organizations respond in fundamentally different ways. Nearly half, 45%, report that issues are resolved within technical teams, often without broader visibility, while only 32% treat these moments as opportunities for transparent review and organizational learning.

These responses reflect a broader leadership choice between containment and learning.

Organizations that respond transparently are better positioned to identify root causes, strengthen systems and build confidence in AI over time. Approaches that emphasise transparency and learning may help strengthen confidence, while approaches focused primarily on containment can limit opportunities for broader organisational learning.

Trust is recognised as strategic but rarely measured as such

A majority of organizations now recognize trust as a strategic asset, with 60 percent classifying it as either a strategic differentiator or a core competitive advantage.

Yet only 28 percent report measuring operational or revenue outcomes linked to trusted AI.

This gap suggests that while trust is increasingly viewed as essential to enterprise performance, it is not yet consistently managed as a business capability. Many organizations still rely on qualitative indicators, proxy measures or no formal measurement approach at all.

The opportunity now is to move beyond recognising trust as important and begin managing it with the same rigor applied to other enterprise capabilities.

Redesigning work

The operating model moves to the foreground

The operating model has moved from the background to the forefront of enterprise transformation.

For years, organizations focused on strategy, technology and capability building as the primary drivers of change. Increasingly, however, leaders are discovering that transformation outcomes are shaped just as much by how the enterprise is organized to make decisions, allocate resources and adapt over time.

Many of the barriers to transformation no longer stem from a lack of ambition, investment or technology. They emerge from the way work is structured, authority is distributed and priorities are translated into action across the organization.

As change becomes more continuous, these underlying mechanisms can either accelerate progress or create friction that limits impact.

This elevates the operating model from a management consideration to a strategic one. The question is no longer whether organizations are transforming, but whether they are organized to sustain transformation over time.

Existing operating models were designed for stability

Most enterprises continue to operate through structures designed for stability rather than continuous execution.

Work remains organized within functions, decisions are coordinated sequentially and processes are optimized locally rather than across end-to-end value streams.

As transformation scales across AI, automation and interconnected workflows, these operating models can introduce increasing friction, slow execution and fragment accountability.

These limitations are most visible in how quickly organizations can translate priorities into action. Only 12 percent report being able to launch new initiatives within three months, while more than half require six months or longer.

In most organizations, priorities are clear. The constraint lies in the ability to translate intent into integrated execution.

The problem is especially visible in environments with complex dependencies such as supply chain, where aligning systems, partners and operations can stretch execution timelines further.

Agility is valued but not structurally enabled

Leaders recognise operational agility as a competitive advantage, with 87 percent reporting that it improves enterprise performance and responsiveness.

This highlights a disconnect: agility is widely prioritized, but many organizations have not designed their operating models to support it.

Decision rights remain fragmented and execution is not consistently aligned across the enterprise.

As a result, organizations experience pockets of speed rather than sustained enterprise responsiveness. Work may move quickly within individual functions or initiatives, but coordination across the enterprise remains inconsistent, limiting the ability to translate priorities into action at scale.

KPMG’s framing here is useful: transformation is rarely constrained by ambition or technical possibilities. It is typically constrained by how work moves across the enterprise.

Fragmented operating models can create isolated improvements but struggle to deliver coordinated enterprise performance.

No single constraint explains AI scale

Organizations are managing interdependent constraints across technology, data, talent and governance simultaneously.

No single issue dominates.

The challenge lies in coordinating systems, capabilities and decisions in ways that allow transformation to scale across functions.

Responsibility for embedding AI often remains distributed across functions rather than owned at enterprise level. Without clear ownership, organizations can struggle to align priorities and coordinate execution.

Twenty nine percent identify implementation cost and unclear ROI as their biggest AI constraint, 23% identify technical and data infrastructure, 22% talent shortages, 16% trust and 10% AI literacy.

The spread itself is the point: enterprise AI is constrained by a system of interconnected issues rather than one technology bottleneck.

Execution now spans connected ecosystems

As operating models become more complex, work is increasingly distributed across internal teams, external partners and interconnected platforms.

Organizations are increasingly extending their capabilities through partners, platforms and managed services. This can accelerate execution, but it also changes where performance is created.

Success is no longer determined solely by how effectively organizations operate internally. It depends on how effectively work moves across an extended ecosystem.

Expanding capability without improving coordination can make it difficult to increase performance. Organizations that fail to integrate work across internal teams, external partners and technology platforms often create more friction than value.

The next frontier is human AI coordination

More than one-third of leaders, 34%, identify the greatest opportunity in combining AI agents with human decision-making, compared with 25% for AI alone and 20% for traditional automation.

This signals a shift in how performance is created.

As work becomes increasingly distributed across teams, partners and platforms, organizations should rethink how execution is coordinated. Realizing value requires more than AI adoption alone. It requires redesigning workflows, decision-making and accountability so people and AI can operate together effectively across functions and systems.

In many organizations, AI is still layered onto existing ways of working rather than used to redesign them. As a result, gains remain incremental and do not consistently translate into enterprise-wide transformation.

The human AI enterprise

Redefining how work is performed

KPMG expects the next phase of enterprise performance to be shaped by how decisively organizations redesign work itself.

As human and machine capabilities increasingly operate together, leading enterprises are moving beyond task-level augmentation toward redesigning how work is coordinated, decisions are made and outcomes are delivered.

In this model, work is organized around end-to-end value streams rather than functions.

AI supports analysis, coordination and execution, allowing people to focus on judgment, oversight and innovation.

This enables work to move more effectively at an enterprise level and connect decisions more directly to value creation. The shift helps improve both efficiency and responsiveness, enabling faster decisions closer to where value is created.

Integrating AI across roles also creates new opportunities for scale, innovation and differentiation.

AI is spreading faster than work is being redesigned

Most organizations are still embedding AI into existing workflows rather than redesigning how work operates.

Tasks are being augmented and processes accelerated, but roles, decision-making and accountability models remain largely unchanged.

This reflects the absence of a true human-AI workforce model — one where work is intentionally redesigned across people, intelligent agents, automation and partners to operate cohesively.

Agentic AI is no longer concentrated in technical domains. While deployment remains highest in technology and IT at 59%, it is now being adopted across core business functions, including marketing and sales at 43%, operations at 42% and risk and compliance at 37%.

This expansion raises an important challenge.

Agentic AI is designed to participate more directly in workflows, decision-making and execution than earlier generations of AI.

As deployment expands, organizations must move beyond adoption and rethink how work is designed.

Without corresponding changes to roles, workflows and accountability, agentic AI may improve individual tasks while leaving broader ways of working largely unchanged.

AI is changing faster than workforce models can adapt

At the heart of many AI initiatives is a fundamental mismatch that many enterprises have not yet addressed.

Most workforce models remain designed around deterministic work, where processes are predefined, decisions follow explicit rules and outcomes are expected to be predictable.

People, supported by enterprise platforms such as ERP, CRM and workflow tools, are organized to deliver consistency at scale.

AI operates differently.

It is inherently probabilistic, producing outputs based on context, patterns and inference rather than fixed instructions.

AI systems can adapt continuously, learn from new data and generate recommendations rather than definitive answers.

While these capabilities are well suited to dynamic environments, they can be difficult to integrate into workforce models designed for predictability and control.

Intelligence is arriving faster than organisations can act on it

The mismatch is increasingly visible in day-to-day execution.

Organizations are deploying technology capable of sensing change in real time, yet execution often still depends on manual coordination and multiple layers of approval.

Intelligent insights sit in dashboards, waiting for action, while opportunities to intervene earlier, prevent issues or dynamically reprioritize work are frequently missed because operating models cannot absorb intelligence at speed.

The challenge is reflected in workforce proficiency.

Despite accelerating investment and deployment, only 19 percent of organizations report a highly proficient workforce in using AI tools.

Improving proficiency may require more than training alone. It may also require rethinking how work is organized, how decisions are made and how accountability is structured so people and AI can operate together more effectively.

Human AI workforce design goes beyond new tools

As AI becomes more deeply embedded into enterprise operations, organizations are beginning to confront a broader workforce challenge: how to organize work when people and intelligent systems contribute side by side.

A true human-AI workforce is not simply about introducing new tools. It requires organizations to rethink how work is assigned, supervised and improved when people and intelligent systems increasingly contribute to shared outcomes.

When asked about the extent to which their organizations support human-AI integration, most leaders report positive progress.

Seventy-five percent agree that leadership promotes AI-human integration, while 72 percent agree that clear roles exist for human oversight and accountability.

However, far fewer express the highest level of confidence that these capabilities are fully embedded within their organizations.

Across leadership support, governance, workforce development and human oversight, a consistent gap exists between agreement and strong agreement.

This suggests that many organizations have begun building the foundations for human-AI integration but have not yet embedded those capabilities into day-to-day operations.

Management becomes orchestration

The emergence of a human-AI workforce changes the role of management itself.

Management shifts from coordinating individuals to orchestrating a broader workforce of human and digital contributors.

The challenge is no longer simply determining where AI should be deployed, but how work should be organized when outcomes are increasingly produced through combinations of human judgment, automation and intelligent systems.

That means performance systems, incentives and career paths designed for a human-only workforce also need to change.

Workday’s perspective in the report makes this explicit: organisations are applying familiar approaches to AI adoption such as change management, training and rollout as though AI were simply another layer of technology.

In practice, the shift is more fundamental.

As AI begins to take on tasks and decisions, work itself starts to fragment. Roles that once operated as coherent units break down into components that should be redistributed across human and machine contributions.

Most organizations have no operating model for managing this shift.

Their systems for performance, incentives and career progression were designed for a human-only workforce, not one where work is increasingly shared across people, AI agents and automation.

Productivity without direction can dissipate

AI is unlocking meaningful capacity, often in the range of 15% to 25%, but organisations are rarely explicit about how that capacity should be used.

What work is deprioritized? What new work is created? How is value measured?

Without clear answers, gains often dissipate or create uncertainty.

The report also highlights roles that no longer map cleanly to the work. What individuals were hired, trained and rewarded to do is increasingly disconnected from where value now sits.

The highest-value work is often more strategic, less defined and not yet reflected in role design or capability models.

Incentives lag behind the shift as well. Employees are being asked to reimagine how work is done but without aligned goals, incentives or performance measures, making the transition uneven and at times resisted.

The conclusion is concise: most organizations are still thinking about AI as a technology deployment challenge. In reality, it is a work redesign challenge.

Rethinking the enterprise

Orchestration becomes a defining capability

The intelligent economy is exposing the limits of traditional transformation models.

Leaders were remarkably consistent in identifying the capabilities that will matter most for future competitiveness.

Across technology and data, customer engagement, operating capabilities and predictive capabilities, a majority view these areas as critical to future advantage.

Yet only a small minority believe they have achieved leadership in delivering them today.

This gap between strategic ambition and organizational capability is becoming one of the defining challenges of enterprise transformation.

As change accelerates, competitive advantage increasingly depends not only on knowing what capabilities matter, but on building the organizational capacity to develop, integrate and scale them throughout the organization.

This requires more than transformation. It requires orchestration: the ability to connect technology, talent, operating and customer capabilities so they reinforce one another rather than compete for attention and investment.

Widespread transformation is not the same as performance

Only 14 percent of organizations consider themselves top performers relative to their peers despite widespread investment in transformation.

Organizations pursuing transformation, particularly in digital and technology domains, are more likely to outperform peers across key business outcomes including innovation, investment capacity and stakeholder trust.

Risk transformation shows an even stronger relationship with performance despite being undertaken by fewer organizations.

But transformation type alone does not determine outcomes.

Organizations that outperform their peers are more likely to connect transformation across operating models, governance, workforce and execution rather than pursuing disconnected initiatives in isolation.

As transformation expands, performance increasingly depends on the ability to coordinate priorities, integrate capabilities and direct execution across the organization.

This calls for more than transformation activity.

It calls for orchestration.

Transformation governance is becoming more complex

Organizations are pursuing transformation across multiple dimensions simultaneously, placing significant pressure on enterprise transformation governance.

Technology transformation is now widespread, with 70 percent of organizations reporting active initiatives.

At the same time, operating model transformation at 51% and enterprise-wide change across systems, processes and people at 48% are underway across a significant share of organizations.

As transformation expands, organizations are managing increasing interdependencies across technology, workforce, operations, governance and risk.

The challenge is no longer initiating change.

It is coordinating execution across interconnected systems, functions and priorities.

Orchestration aligns concurrent transformations

As organizations pursue multiple transformations simultaneously, the challenge shifts from initiating change to ensuring different transformation efforts work together toward common outcomes.

Orchestration provides the mechanism for doing so.

It creates visibility across transformation activity, connects decisions across functions and helps organizations coordinate execution across increasingly interconnected operating environments.

Organizations embedding AI into enterprise strategy and operating models are more likely to realize enterprise-wide value when capabilities are aligned and integrated.

Without this alignment, AI often improves individual activities while broader transformation outcomes remain difficult to scale.

Success in orchestration depends on a set of underlying conditions: strategic alignment, leadership direction, integrated governance, talent depth and strong foundational capabilities.

When these capabilities evolve unevenly, it becomes harder to reconcile transformation success against value and desired outcomes.

Orchestration turns transformation into enterprise performance

Orchestration is not separate from transformation. It is how transformation is directed and sustained.

As transformation scales across functions, systems and workflows, coordination can no longer remain episodic or program-based.

Organizations require continuous alignment across strategy, operations, technology, workforce and governance.

Performance increasingly depends on how effectively organizations integrate these capabilities into coordinated enterprise execution.

Leading organizations outperform not because they undertake more transformation, but because they connect technology, operations, workforce and governance efforts into a coordinated enterprise agenda.

This creates a more direct path from transformation activity to enterprise performance.

Performance is therefore no longer defined by how much transformation an organisation undertakes, but by how effectively it directs it.

Leading organizations orchestrate transformation, making deliberate decisions about where to accelerate, align or stop activity to maximise enterprise value.

Principles of enterprise orchestration

As markets, technologies, operating models and customer expectations evolve simultaneously, organizations must coordinate multiple dimensions of change at once — not sequentially or in isolation.

Orchestration aligns strategy, business models, capabilities, governance and execution into a coordinated enterprise system.

Rather than managing disconnected initiatives, leading organizations continuously align priorities, direct execution and adapt dynamically as conditions change.

Effective orchestration follows several principles.

Organisations should align strategic priorities, business model choices and value streams around measurable value outcomes. Experiences, capabilities across humans, AI and systems, and operating models should be designed as an integrated whole.

Real-time signals from performance, customers, markets and risk should be used to adjust direction dynamically.

Enterprise-level ownership should resolve trade-offs across growth, risk, speed and efficiency.

In contrast, KPMG warns against redefining strategy without reshaping the operating model that must deliver it, optimising value streams or experiences in isolation from enterprise priorities, treating AI, digital or operating change as standalone transformations and governing change through disconnected forums that delay decisions and dilute accountability.

Effective orchestration depends on shared definitions of value, integrated governance, enterprise-wide visibility and coordinated decision-making across functions and systems.

The report sums it up simply: orchestration is not about alignment alone. It is about making the right calls — where to scale, where to stop and how to turn AI into better customer outcomes.

Leadership decisions make orchestration visible

Enterprise orchestration becomes visible in moments of trade-off.

When growth and risk collide, leaders make speed conditional on trust. AI initiatives scale only when governance, controls and accountability are embedded from the outset. Risk is not reviewed after execution; it determines whether execution can scale.

When everything looks successful, leaders introduce explicit stopping rules. Local success alone is not enough. Initiatives must integrate and reinforce shared capabilities rather than perform in isolation.

When strategy stalls in execution, leaders redesign decision rights around enterprise outcomes rather than functional ownership. Accountability is distributed across interconnected teams with authority to act closer to execution.

When governance cannot keep pace, leaders embed governance directly into execution. Real-time thresholds trigger escalation, intervention or override rather than relying on periodic review cycles.

When resources are trapped in the past, leaders continuously reallocate capital and talent toward enterprise priorities rather than fixed annual plans or functional ownership structures.

These examples make the difference between transformation management and orchestration concrete. Orchestration actively determines what can move, who can decide, when intervention occurs and where scarce resources should go.

Customer experience becomes a real time test of orchestration

As AI becomes embedded across decisions and workflows, customer experience increasingly reveals whether the enterprise is truly integrated.

Every interaction draws on multiple systems, data sources, workflows and decision rights.

Where orchestration is weak, experiences fragment. Customers repeat themselves, issues move slowly across functions and intelligence fails to translate into action.

Where orchestration is strong, experiences become faster, more adaptive and more predictive.

Customer experience therefore becomes a real-time indicator of how effectively the enterprise aligns data, decision-making, trust and execution across the organization.

KPMG turns this into a useful board level test.

Are data, platforms and AI capabilities integrated enough to scale intelligence, or is the enterprise still layering new capabilities onto fragmented foundations?

Is governance embedded into execution strongly enough to accelerate AI adoption with confidence, or is risk still being applied after the fact?

Where do handoffs, delays and functional boundaries still prevent strategy from translating into coordinated execution?

Have roles, workflows and decision-making been redesigned for human-AI collaboration, or is AI only being added to existing work?

And who has the mandate to resolve trade-offs, stop disconnected initiatives and reallocate resources toward enterprise priorities?

What transformation leaders do differently

Transformation leaders are distinguished less by the scale of their ambition than by the choices they make: what they prioritize, how they sequence change and how they align and direct the enterprise to deliver it.

The first choice is to prioritize adaptability over efficiency.

Most organizations continue to pursue performance through process efficiency. Leading organizations make a different choice by prioritizing workforce adaptability as a primary driver of execution.

They are less likely to prioritise process optimisation, at 39% compared with 55% among other organisations. They place greater emphasis on reskilling, employee experience and the ability to continuously reconfigure how work gets done.

The second is to prioritize enterprise alignment over transformation volume.

Most organizations respond to pressure by increasing the volume of transformation activity. Leaders instead build enterprise orchestration capabilities that enable the organization to adapt more effectively, coordinate execution at scale and compound value over time.

They are nearly nine times more likely to be leading in enterprise capability maturity and three times more likely to significantly outperform competitors.

The third is to align the enterprise before scaling AI.

Most organizations are attempting to scale AI technology within fragmented structures. Leading organizations do the opposite, integrating the enterprise first and then scaling technology across the connected system.

They drive stronger alignment across strategy, operating model, talent and technology and demonstrate a greater ability to translate AI adoption into measurable business outcomes by integrating execution across the enterprise.

The fourth is to design trust and risk as foundations that enable speed.

Most organizations treat risk and governance as constraints on transformation. Leading organizations design them as enablers of speed by embedding trust directly into how the enterprise operates.

Leading organisations are more likely to adopt proactive AI risk management embedded across strategy and the technology lifecycle, and more than twice as likely to treat trust as a source of competitive differentiation, at 70% versus 31%.

The fifth is to leverage external partners to accelerate progress.

Most organisations delay transformation until internal capabilities are in place. Leading organisations use ecosystem partnerships, alliances and platforms to accelerate execution and extend capabilities as conditions change.

Leaders are twice as likely to leverage managed services and are more likely to integrate ecosystem partners directly into how value is created, delivered and scaled.

Alliance perspectives

Work is being redefined faster than organisations can adapt

The Workday perspective in the report focuses directly on the operating model implications of AI.

Organizations are applying familiar approaches to AI adoption — change management, training and rollout — as though this were simply another layer of technology.

In practice, the shift is more fundamental.

Most organizations have no operating model for managing this shift.

Their systems for performance, incentives and career progression were designed for a human-only workforce, not one where work is increasingly shared across people, AI agents and automation.

AI is unlocking meaningful capacity, often in the range of 15% to 25%, but organisations are rarely explicit about how that capacity should be used.

What work is deprioritised? What new work is created? How is value measured?

Without clear answers, gains often dissipate or create uncertainty.

The report also argues that roles no longer map neatly onto the work. What individuals were hired, trained and rewarded to do is increasingly disconnected from where value now sits. The highest-value work is often more strategic, less defined and not yet reflected in role design or capability models.

Employees are simultaneously being asked to reimagine how work is done without aligned goals, incentives or performance measures, making the transition uneven and at times resisted.

Most organizations have mature, decades-old systems for managing human work that track output, performance, skills and progression. They have almost no equivalent for managing machines as contributors.

As a result, AI is often governed as a tool even when it is actively participating in how work gets done.

The conclusion is clear: organisations moving ahead treat AI as a work design challenge rather than simply an adoption challenge. They redefine roles, make explicit decisions about how capacity is redeployed and build new models for coordinating human and machine contribution as one workforce.

Finance moves from reporting performance to continuously steering it

The Oracle perspective describes a similar shift in finance.

Finance transformation has historically focused on efficiency: closing faster, automating manual work and standardizing processes.

These remain essential but are no longer sufficient in an environment defined by volatility, margin pressure and rising expectations from leadership.

The role of finance is being reshaped by a more fundamental expectation: helping the business make better decisions faster and with greater confidence.

This is changing the operating rhythm of the function.

Rather than operating in cycles — close, report, forecast — finance is increasingly expected to operate continuously, integrating financial, operational and external data to deliver real-time insight into performance, risk and opportunity.

Variance analysis and forecasting therefore shift from explaining past performance towards identifying what is likely to happen and what actions should be taken in response.

Static forecasts give way to rolling, scenario-based models that evolve in real time as new data becomes available.

Finance increasingly integrates data across demand, supply, cost, customer behaviour and workforce to create a more complete view of business performance.

But most organisations are not structurally prepared for the shift. Data remains fragmented, processes are inconsistently customised and finance teams continue to spend disproportionate time preparing data rather than interpreting it and guiding decisions.

The intended end state is finance as a central decision layer, continuously assessing performance, adjusting forecasts and guiding actions across the enterprise.

The question changes from “what happened?” to “what does this mean, and what should we do next?”

Supply chain visibility has outpaced the operating model

The SAP perspective makes the same argument through supply chain.

Technology has advanced to the point where end-to-end visibility and coordination are increasingly possible.

Yet in practice, most supply chains remain fragmented, not just across the enterprise but within it. Planning, manufacturing and logistics continue to operate as distinct domains, each with its own data, technologies and decision-making cadence.

The constraint is now less technical than organisational.

Even in advanced organisations, there is often no clear ownership model for decisions that span functions, let alone extend across finance, suppliers and partners.

As a result, transformation efforts frequently default to technology upgrades or functional optimisation without addressing the operating structures required to manage the supply chain end-to-end.

Organisations can increasingly see across the supply chain but lack the structures and incentives to act on the insight in a coordinated way.

Many initiatives begin with system modernisation rather than a clearly defined business objective, creating progress that can be difficult to measure and sustain.

Even highly digitised organisations often revert to analogue processes once they move beyond the enterprise boundary, limiting the ability to coordinate suppliers, partners and customers in real time.

The lesson is that technology has made real-time supply chain visibility increasingly possible, but the operating model required to act on that visibility still lags behind.

Risk is still moving at human speed

The ServiceNow perspective addresses the same mismatch through risk.

Across organisations, risk is not experienced uniformly. Cyber and technology teams already operate in a world of continuous monitoring and rapid response.

Elsewhere — across legal, operations and enterprise risk — risk is still often managed episodically, assessed at intervals and governed within functional boundaries.

The gap between those models becomes increasingly difficult to sustain.

As AI scales, the pace at which risk emerges and spreads is compressing.

What begins as a vulnerability in one domain — cyber, third party or data — can now propagate across platforms, processes and physical environments within minutes.

The relationship between cause and consequence is no longer linear, and the time to impact is no longer measured in human cycles.

As agents interact with agents, decisions and actions occur at machine speed. Risk events no longer unfold gradually — they cascade.

Most organisations have not fully internalised what that compression of time means for exposure.

Continuous monitoring exists in parts of the enterprise, particularly technology, but has not extended across the full set of business processes where AI is now embedded.

KPMG’s contributor argues that risk only begins to scale with the business when it becomes an underlying operating fabric connecting technology, processes, data and third parties.

In that model, risk does not slow the business. It defines how far and how fast the business can move.

Sector and function highlights

Different sectors show different transformation constraints

KPMG finds substantial variation by sector because operating models, regulation and the integration of technology, data and decisions differ materially.

In banking and capital markets, transformation is strongest in control functions but constrained at enterprise level. AI is deeply embedded in risk, compliance and operational efficiency, including fraud detection, credit underwriting and regulatory reporting. Momentum is weaker in customer facing innovation and new business models.

In industrial manufacturing, transformation is delivering operational gains but not yet strategic advantage. AI is widely used in predictive maintenance, production planning and quality control, but is less mature in innovation, new revenue streams and strategic decision making.

In consumer markets, retail and leisure, AI is advancing quickly in customer engagement but has not yet been integrated into the wider enterprise layer. Front office use in personalisation and pricing is more advanced than the integration of customer intelligence into operations, supply chain and decision systems.

In energy, natural resources and chemicals, AI remains concentrated in targeted high value use cases such as asset optimisation, demand forecasting and field operations.

In healthcare, transformation is progressing but constrained by trust, regulation and system complexity. AI is increasingly used in diagnostics, patient triage and capacity planning, but scaling is limited by fragmented operating environments.

In technology and telecommunications, AI is more widely embedded and supports both performance and growth. The issue is shifting away from initial adoption towards sustaining advantage through continual integration, coordination and innovation.

In government and the public sector, transformation activity is high but enterprise impact remains difficult to scale. Legacy complexity, fragmented delivery and the pace of institutional change constrain the translation of AI adoption into sustained value.

HR is moving towards total workforce management

KPMG describes HR as having earned a greater strategic mandate but not yet being structured to support the shift towards total workforce management.

HR is playing an increasingly central role in enterprise transformation, with stronger alignment to business outcomes and growing ownership of productivity and performance.

This is accelerating a move towards a workforce model integrating employees, contingent talent and AI.

But progress remains early.

Sixty one percent of organisations report evolving toward this model, while only 11 percent have fully transitioned.

Most still operate through role based structures, siloed workforce planning and misaligned talent systems.

HR is therefore becoming more strategic in talent management while the underlying architecture of work remains largely human centric.

Risk is aligned strategically but not operationally

Risk is increasingly positioned as a strategic enabler, with 84 percent of organisations reporting alignment between risk management and business strategy.

However, the strategic shift has not been matched operationally.

In many organisations, risk still functions primarily as a control mechanism applied after decisions are made rather than shaping how decisions are defined, evaluated and governed.

This gap is particularly visible in AI.

Adoption is strongest in structured, quantitative areas such as risk assessment and cybersecurity but remains more limited in judgement intensive areas such as model governance, ethical decision making and enterprise risk trade offs.

As a result, risk remains unevenly integrated into the workflows, technologies and decisions it is meant to govern.

Finance is becoming an enterprise decision engine, but slowly

Survey findings show AI already widely embedded across finance, with 94 percent of organizations reporting at least partial adoption and more than 75 percent seeing value creation and or reduced manual effort across reporting, forecasting and analysis.

Finance is also emerging as a central node in enterprise planning, with some of the highest levels of AI integration and expected future impact across functions.

However, most organizations continue to focus on optimizing existing processes rather than reconfiguring finance as a forward-looking, decision-oriented function.

Finance often remains anchored in periodic reporting cycles, backward-looking KPIs and fragmented planning rather than enabling continuous, real-time insight and decision support.

System integration challenges, skills gaps and unclear ROI continue to limit the ability to scale end-to-end automation and advanced analytics.

Leading in the next phase of transformation

Leading in the next phase of transformation calls for organizations to adopt a new set of strategic priorities for how the enterprise is designed, operated and continuously evolved.

The first is to rebuild technology foundations for the intelligence layer.

Organizations that scale AI successfully tend to invest first in modern technology foundations. They often move beyond fragmented modernization to build resilient, interoperable platforms where data, technologies and AI capabilities operate as a connected system.

Integration can be designed in from the outset to help enable AI to scale safely, efficiently and continuously.

The second is to establish trust foundations for speed and value.

Most leading organizations typically embed governance, responsible AI, safety and security directly into decision-making and execution rather than layering them on after deployment.

By integrating risk management, controls and accountability into operating models and architectures, trust can become an enabler of speed, adoption and scalable AI execution.

The third is to redesign the operating model around end-to-end value streams.

Most high-performing organizations structure work around how value is created and delivered rather than around functional boundaries.

Leaders redesign operating models so strategy, customers, operations and technology operate as a connected execution system.

This can reduce handoffs, shorten decision cycles and enable AI to create impact across the full flow of value.

The fourth is to reimagine work for human-AI collaboration.

Most leading organizations make workforce adaptability a core operating capability.

Rather than relying on episodic upskilling, they redesign roles and workflows so human expertise and AI operate together by design.

Learning is embedded into the flow of work, capabilities evolve continuously and the workforce becomes a driver of AI scale rather than a constraint.

The fifth is to rethink enterprise transformation as agile enterprise orchestration.

Most leaders treat transformation as a coordinated enterprise capability rather than a collection of disconnected initiatives.

They orchestrate strategy, execution and investment across functions, helping ensure capabilities reinforce one another over time.

Leading organisations increasingly model enterprise value streams, decisions, risks and controls in ways that both humans and AI can understand, supporting more coordinated execution at scale.

Conclusion

KPMG’s Transforming the Enterprise 2026 makes a distinction that runs through the entire report: transformation activity is not the same thing as transformation performance.

Most large organisations are already transforming. The average organisation in the research is managing 3.5 transformations simultaneously. AI is spreading across functions, productivity is improving and leaders understand which capabilities matter.

Yet only 14% consider themselves top performers.

The gap is increasingly explained by how the organisation itself operates.

AI can automate a task without changing the workflow. A function can become more productive without improving enterprise performance. A transformation programme can succeed locally while increasing fragmentation elsewhere.

The next stage therefore requires organisations to connect what they have previously managed separately.

Technology and data form an intelligence layer. Trust and governance become part of execution rather than downstream review. Operating models move towards end to end value streams. Human and AI roles are intentionally redesigned. Transformation itself becomes continuous enterprise orchestration.

The human AI element is particularly important.

Agentic AI is already spreading into business functions faster than work is being redesigned. Only 19% of organisations say their workforce is highly proficient, and even organisations that promote human AI integration often have not fully embedded the necessary roles, accountability, training and governance into daily operations.

KPMG therefore does not treat workforce adaptation as simply a training problem.

It is a work design problem.

Roles, workflows, decision rights, performance systems, incentives and management structures all need to evolve as intelligent systems become active contributors to work.

The same principle applies to governance. As decision making moves towards machine speed, governance cannot continue to operate entirely through periodic review and human cycles. Trust, risk thresholds, accountability, escalation and intervention increasingly need to exist inside the operating system of the enterprise itself.

This brings the argument back to orchestration.

The most capable organisations will not necessarily be those running the largest number of transformations or deploying the most AI. They will be the organisations that can make deliberate choices about what to scale, what to stop, where humans should remain accountable, where AI should act and how enterprise resources should move as conditions change.

The next transformation advantage therefore comes less from initiating more change and more from building an organisation capable of continuously directing change.

Key takeaways

Transformation has become continuous and enterprise wide, but enterprise structures have not necessarily evolved at the same pace. Ninety four percent of organisations are running at least two transformations simultaneously, yet only 14% consider themselves top performers.

AI makes the gap more visible. Organisations are already seeing productivity improvements, but AI deployment is often moving faster than operating model change. Local gains therefore do not automatically compound into enterprise performance.

The intelligence layer is becoming foundational. AI needs integrated data and technology, but it also needs shared enterprise context, decision logic, governance and business ontology if it is to operate coherently across functions and systems.

Trust is moving from a downstream safeguard into an operational capability. As intelligent systems become more autonomous, governance needs to sit inside decision making and execution, with real time thresholds, intervention rights and accountability.

Work needs to be redesigned rather than simply augmented. Agentic AI is already spreading across technology, sales, operations, risk, finance and HR, but most organisations still retain workforce models built around human only, deterministic processes.

The next productivity frontier is therefore human AI coordination. KPMG finds that leaders expect greater future performance from AI agents combined with human decision making than from AI agents alone.

Management changes with this model. Managers increasingly orchestrate combinations of employees, agents, automation, platforms and external partners rather than simply coordinating people.

Transformation leadership also changes. The relevant question is no longer how many initiatives the organisation can launch. It is whether leadership can establish the decision rights, stopping rules, resource allocation mechanisms and real time visibility required to direct those initiatives as one system.

KPMG’s central argument can therefore be reduced to one line: the challenge is no longer initiating transformation. It is orchestrating it.

Download the report

Abinesh Kanagaraj

KPMG: Transforming the Enterprise 2026

Key takeaways

Executive summary

Key stats and quotable claims

Overall summary

Deep dive

Enterprise transformation has entered a new phase

Early AI gains expose the next transformation challenge

The gap between activity and performance

Autonomous systems are putting governance under strain

Coordination is not orchestration

Three conditions for scaling transformation

Five strategic priorities

Enterprise transformation in the age of AI

Transformation is becoming continuous and enterprise wide

Misalignment becomes a primary constraint

Enterprise orchestration offers a path forward

Three pathways towards AI performance

The enterprise transformation agenda is an integrated capability system

Rebuilding the foundations

Why AI can fail to scale

Hybrid ecosystems increase complexity

The intelligence layer becomes pivotal

Productivity remains the dominant measure of AI value

The dominant architecture is becoming hybrid

Existing systems are being extended rather than replaced

Data alone is not enough

Shared enterprise context requires ontology

Trust: a foundation for scaling transformation

Trust is becoming operational

Most organisations have not embedded trust

Leadership behaviour determines whether AI failures become learning

Trust is recognised as strategic but rarely measured as such

Redesigning work

The operating model moves to the foreground

Existing operating models were designed for stability

Agility is valued but not structurally enabled

No single constraint explains AI scale

Execution now spans connected ecosystems

The next frontier is human AI coordination

The human AI enterprise

Redefining how work is performed

AI is spreading faster than work is being redesigned

AI is changing faster than workforce models can adapt

Intelligence is arriving faster than organisations can act on it

Human AI workforce design goes beyond new tools

Management becomes orchestration

Productivity without direction can dissipate

Rethinking the enterprise

Orchestration becomes a defining capability

Widespread transformation is not the same as performance

Transformation governance is becoming more complex

Orchestration aligns concurrent transformations

Orchestration turns transformation into enterprise performance

Principles of enterprise orchestration

Leadership decisions make orchestration visible

Customer experience becomes a real time test of orchestration

What transformation leaders do differently

Alliance perspectives

Work is being redefined faster than organisations can adapt

Finance moves from reporting performance to continuously steering it

Supply chain visibility has outpaced the operating model

Risk is still moving at human speed

Sector and function highlights

Different sectors show different transformation constraints

HR is moving towards total workforce management

Risk is aligned strategically but not operationally

Finance is becoming an enterprise decision engine, but slowly

Leading in the next phase of transformation

Conclusion

Key takeaways

KPMG: 2026 Frontiers Report - AI as the Frontier Multiplier

Capgemini: The Multi Year AI Advantage