Brief

SAP Sapphire 2026: ERP Moves Beyond Transactions to Orchestration

SAP positions its core products as moving from systems of record to systems of context, integrating the enterprise with AI and agents.

Florian Braun, Dominic Wheeler, Donna Enverga, Ignacio Gómez Garcia-Onieva, and Juan Francisco (Paco) Jimenez

First published on мая 27, 2026
min read


Резюме Some are arguing that ERP software is less necessary in the age of AI, but SAP laid out its argument for the enterprise resource planning layer to remain an essential orchestrator of activity. SAP sees its software moving from a system of record to a system of context—managing enterprise transactions and data and bridging systems. Modernization of these systems is more necessary than ever, and it should happen in coordination with the deployment of AI and agents. ERP transformation is increasingly the foundation on which enterprise AI strategies either succeed or struggle to scale. SAP set out its vision for the Autonomous Enterprise at two recent Sapphire events in Orlando and Madrid, where the company defined the elements of a more connected enterprise platform in which data, workflows, and AI operate together rather than as separate layers. SAP introduced a unified Business AI Platform, positioned Business Data Cloud as a critical foundation for trusted enterprise data and business context, placed Joule at the center of the user and agent experience, and emphasized the role of SAP Knowledge Graph in helping AI understand how business data, processes, and decisions connect. SAP also highlighted a broader ecosystem of partnerships, including OpenAI, Anthropic, Mistral, and Palantir, to strengthen the platform around its core enterprise applications. Whether this is truly a new beginning rather than a continuation of SAP’s long-term roadmap remains open to interpretation. But one message was clearer than ever: SAP increasingly sees the future value of ERP not only in transaction processing but in owning the enterprise context and orchestration layer around it. The strategic battleground is moving above the ERP core Some AI-native companies argue that ERP might no longer be necessary, given agents’ ability to manage more complex processes. Our view is that the ERP backbone remains essential: We still need a layer where transactions are recorded, auditable, and traceable—without autonomous changes from agents or hallucinations. If anything, AI increases that need; it doesn’t reduce it. But value is moving above the core system itself. The strategic question is no longer just who owns the transaction engine; it’s who owns the data, business context, and orchestration layer that AI systems depend on. SAP’s message at Sapphire was that ERP is evolving from a system of record into a system of context. AI agents cannot operate effectively from isolated systems alone. They need access to connected enterprise data, process logic, workflows, and governance across ERP, human resources, CRM, supply chain, and external platforms. This explains SAP’s focus on Business Data Cloud, Knowledge Graph, Joule agent orchestration, and the wider Business AI Platform. The aim is not simply to store enterprise data but to connect and contextualize it so AI can act on it more reliably. A clearer architecture stack is emerging: Systems of record manage core enterprise transactions. A data and orchestration layer connects business context across systems. AI agents and analytics operate on top of that governed context. For CIOs, the implications are that ERP modernization, data architecture, governance, and AI strategy can no longer be treated as separate transformation agendas. Organizations that run ERP and AI as discrete workstreams may later find themselves redesigning one to accommodate the other. What CIOs need to consider now Clean core is becoming a route to faster AI value. Clean core is no longer just a technical standard. Simpler systems with fewer customizations are generally cheaper, faster, and easier to maintain. They also create a more stable foundation for data consistency, governance, integration, and future AI adoption. For clients, this means rationalizing applications, reducing unnecessary customization and shadow IT, and consolidating data around a clearer semantic layer. Organizations with cleaner and more standardized environments will likely find it easier to integrate, operationalize, and monetize new AI capabilities over time. Migration paths are becoming more flexible. Our clients’ questions at Sapphire were less about the need to modernize their ERP platforms and more about when and how to move, and how to sequence this modernization along with agentification—creating agents to manage more processes. They wanted to know how to extract value from legacy landscapes and where to leverage AI through the transition. Sapphire suggested a more flexible migration landscape, with SAP and partners introducing AI-supported tools to reduce migration effort, particularly in complex data migration scenarios. This gives organizations more room to sequence the journey: prepare the data foundation, test targeted AI use cases, and move toward modern ERP when the value case, cost, risk, and timing are clear. The destination may be similar, but the path does not need to be one-size-fits-all. AI may reduce delivery effort but not transformation complexity. SAP demonstrated growing use of AI across process mining, configuration, testing, documentation, and migration activities, with SAP stating these capabilities could reduce ERP migration effort by about half. While AI may reduce selected delivery effort, it does not eliminate the need for business-led process redesign, organizational change, integration simplification, and data remediation. Clients continue to question how these savings will be achieved in practice, and conversations with system integrators suggest ways of working will also need to evolve to fully realise them. CIOs should challenge delivery partners to demonstrate where AI will materially reduce cost, time, or risk, and how those gains will be measured. AI access and platform control are becoming more strategic. A fundamental shift is underway in how enterprise software vendors manage AI agent access to enterprise platforms and data. SAP has acted explicitly, requiring approval for certain third-party AI agents to access its systems. For CIOs, this is an area to actively monitor, as decisions around APIs, extensibility, and platform governance may increasingly influence future flexibility and control across enterprise AI ecosystems. As tech vendors race to monetize the AI layer, CIOs need to watch the small print on total operating costs. In April 2026, SAP quietly published a revised API policy requiring SAP endorsement for AI-driven and agentic access patterns—which SAP describes as standardizing protections rather than restricting access. Even so, as enterprise platform vendors move to govern the AI layer above their data, the commercial terms attached to that access are becoming a material consideration for total cost of ownership. As one analyst observed at the conference, platform choices made in 2026 may effectively set the terms of enterprise AI contracts for the next decade. CIOs need to play a more active role as transformation guardians. As ERP modernization and AI strategies become more interconnected, CIOs are increasingly expected to shape architecture decisions and understand their implications. They also need to revisit commercial models to ensure AI investments translate into measurable business value. Sapphire 2026 continued to evolve SAP’s positioning from an ERP vendor to an orchestration platform for autonomous enterprise operations. Compared with last year’s conference, there’s more evidence of operational use cases, ecosystem integration, implementation acceleration, and platform pragmatism. The broader implication for CIOs is that ERP modernization, data architecture, governance, and AI investments can no longer be treated as separate agendas. The path to value still depends on foundational work many organizations have not yet completed: clean data, standardized processes, and clear architecture decisions. Organizations will need to assess where they genuinely sit on the readiness curve and sequence investments accordingly. ERP transformation is increasingly becoming the foundation on which enterprise AI strategies either succeed or struggle to scale. Catherine Hicks, Gaurav Sharma, Radhika Vyas, and Joseph Ravenscroft also contributed to this report.

Резюме

Some are arguing that ERP software is less necessary in the age of AI, but SAP laid out its argument for the enterprise resource planning layer to remain an essential orchestrator of activity.
SAP sees its software moving from a system of record to a system of context—managing enterprise transactions and data and bridging systems.
Modernization of these systems is more necessary than ever, and it should happen in coordination with the deployment of AI and agents.
ERP transformation is increasingly the foundation on which enterprise AI strategies either succeed or struggle to scale.

SAP set out its vision for the Autonomous Enterprise at two recent Sapphire events in Orlando and Madrid, where the company defined the elements of a more connected enterprise platform in which data, workflows, and AI operate together rather than as separate layers.

SAP introduced a unified Business AI Platform, positioned Business Data Cloud as a critical foundation for trusted enterprise data and business context, placed Joule at the center of the user and agent experience, and emphasized the role of SAP Knowledge Graph in helping AI understand how business data, processes, and decisions connect. SAP also highlighted a broader ecosystem of partnerships, including OpenAI, Anthropic, Mistral, and Palantir, to strengthen the platform around its core enterprise applications.

Whether this is truly a new beginning rather than a continuation of SAP’s long-term roadmap remains open to interpretation. But one message was clearer than ever: SAP increasingly sees the future value of ERP not only in transaction processing but in owning the enterprise context and orchestration layer around it.

The strategic battleground is moving above the ERP core

Some AI-native companies argue that ERP might no longer be necessary, given agents’ ability to manage more complex processes. Our view is that the ERP backbone remains essential: We still need a layer where transactions are recorded, auditable, and traceable—without autonomous changes from agents or hallucinations. If anything, AI increases that need; it doesn’t reduce it.

But value is moving above the core system itself. The strategic question is no longer just who owns the transaction engine; it’s who owns the data, business context, and orchestration layer that AI systems depend on.

SAP’s message at Sapphire was that ERP is evolving from a system of record into a system of context. AI agents cannot operate effectively from isolated systems alone. They need access to connected enterprise data, process logic, workflows, and governance across ERP, human resources, CRM, supply chain, and external platforms.

This explains SAP’s focus on Business Data Cloud, Knowledge Graph, Joule agent orchestration, and the wider Business AI Platform. The aim is not simply to store enterprise data but to connect and contextualize it so AI can act on it more reliably.

A clearer architecture stack is emerging:

Systems of record manage core enterprise transactions.
A data and orchestration layer connects business context across systems.
AI agents and analytics operate on top of that governed context.

For CIOs, the implications are that ERP modernization, data architecture, governance, and AI strategy can no longer be treated as separate transformation agendas. Organizations that run ERP and AI as discrete workstreams may later find themselves redesigning one to accommodate the other.

What CIOs need to consider now

Clean core is becoming a route to faster AI value. Clean core is no longer just a technical standard. Simpler systems with fewer customizations are generally cheaper, faster, and easier to maintain. They also create a more stable foundation for data consistency, governance, integration, and future AI adoption.

For clients, this means rationalizing applications, reducing unnecessary customization and shadow IT, and consolidating data around a clearer semantic layer. Organizations with cleaner and more standardized environments will likely find it easier to integrate, operationalize, and monetize new AI capabilities over time.

Migration paths are becoming more flexible. Our clients’ questions at Sapphire were less about the need to modernize their ERP platforms and more about when and how to move, and how to sequence this modernization along with agentification—creating agents to manage more processes. They wanted to know how to extract value from legacy landscapes and where to leverage AI through the transition.

Sapphire suggested a more flexible migration landscape, with SAP and partners introducing AI-supported tools to reduce migration effort, particularly in complex data migration scenarios. This gives organizations more room to sequence the journey: prepare the data foundation, test targeted AI use cases, and move toward modern ERP when the value case, cost, risk, and timing are clear. The destination may be similar, but the path does not need to be one-size-fits-all.

AI may reduce delivery effort but not transformation complexity. SAP demonstrated growing use of AI across process mining, configuration, testing, documentation, and migration activities, with SAP stating these capabilities could reduce ERP migration effort by about half. While AI may reduce selected delivery effort, it does not eliminate the need for business-led process redesign, organizational change, integration simplification, and data remediation. Clients continue to question how these savings will be achieved in practice, and conversations with system integrators suggest ways of working will also need to evolve to fully realise them. CIOs should challenge delivery partners to demonstrate where AI will materially reduce cost, time, or risk, and how those gains will be measured.

AI access and platform control are becoming more strategic. A fundamental shift is underway in how enterprise software vendors manage AI agent access to enterprise platforms and data. SAP has acted explicitly, requiring approval for certain third-party AI agents to access its systems. For CIOs, this is an area to actively monitor, as decisions around APIs, extensibility, and platform governance may increasingly influence future flexibility and control across enterprise AI ecosystems.

As tech vendors race to monetize the AI layer, CIOs need to watch the small print on total operating costs. In April 2026, SAP quietly published a revised API policy requiring SAP endorsement for AI-driven and agentic access patterns—which SAP describes as standardizing protections rather than restricting access. Even so, as enterprise platform vendors move to govern the AI layer above their data, the commercial terms attached to that access are becoming a material consideration for total cost of ownership. As one analyst observed at the conference, platform choices made in 2026 may effectively set the terms of enterprise AI contracts for the next decade.

CIOs need to play a more active role as transformation guardians. As ERP modernization and AI strategies become more interconnected, CIOs are increasingly expected to shape architecture decisions and understand their implications. They also need to revisit commercial models to ensure AI investments translate into measurable business value.

Sapphire 2026 continued to evolve SAP’s positioning from an ERP vendor to an orchestration platform for autonomous enterprise operations. Compared with last year’s conference, there’s more evidence of operational use cases, ecosystem integration, implementation acceleration, and platform pragmatism.

The broader implication for CIOs is that ERP modernization, data architecture, governance, and AI investments can no longer be treated as separate agendas. The path to value still depends on foundational work many organizations have not yet completed: clean data, standardized processes, and clear architecture decisions. Organizations will need to assess where they genuinely sit on the readiness curve and sequence investments accordingly. ERP transformation is increasingly becoming the foundation on which enterprise AI strategies either succeed or struggle to scale.

Catherine Hicks, Gaurav Sharma, Radhika Vyas, and Joseph Ravenscroft also contributed to this report.