As discussed in a new whitepaper, , this perspective is grounded in interviews with supply chain leaders across six industries: automotive electronics and software, agricultural equipment, chemicals, global technology, automotive supply, and home appliances.

Their experiences reveal where companies are investing, where adoption challenges remain, and where the next wave of value is likely to emerge.

Supply chains are entering an era of permanent disruption

Four structural forces are reshaping global supply chains simultaneously: geopolitical instability, economic pressure, demographic shifts, and accelerated digital transformation.

Since 2017, relative to trade among closer partners, signaling growing fragmentation in global commerce. , while labor shortages and digital skill gaps continue to constrain operations.

Europe alone could face by 2028, and 63% of companies cite .

Together, these pressures are pushing supply chains beyond the limits of the traditional “plan-source-make-deliver” model.

Companies are shifting from optimization to AI-enabled orchestration

Supply chains are increasingly viewed as strategic levers for resilience, service differentiation, and competitive advantage.

Across all six companies interviewed, each is investing in at least three forward-looking AI use cases in planning alone.

• A leading agricultural equipment company has deployed more than 1,000 AI agents to support orchestration, scenario planning, and value chain visibility. A global chemicals company is embedding AI across planning and scenario management while emphasizing explainability and trust.
• A home appliance company is applying AI selectively to improve forecasting, transport optimization, warehouse safety, and logistics costs.

The common theme: organizations are redesigning how the enterprise senses, decides, and acts.

Resilience is now defined by decision velocity

In today’s fragmented environment, resilience is no longer about static buffers. It is about how quickly companies can convert disruption signals into coordinated action across sourcing, production, planning, and logistics.

• An automotive electronics and software company centralized electronics ordering across roughly 30 plants and redesigned crisis-management processes, reducing disruption response times by approximately 95%.
• A global technology company adopted a regional “two-leg” supply chain model, using inventory strategically to respond faster to disruptions.

The emerging differentiator is not forecast accuracy alone, but the speed from disruption detection to execution. Visibility remains important, but visibility without coordinated action is no longer enough.

Trust and governance are the biggest barriers to scaling AI

Despite rapid interest, . The challenge is not model accuracy alone; it is trust, explainability, fragmented systems, and manual overrides.

• One global chemicals company found that scaling AI depended less on technical performance and more on whether users could understand and trust the outputs. This led to stronger human-in-the-loop governance and progressive autonomy thresholds.
• A major automotive electronics company requires transparent, traceable AI reasoning before planners rely on AI-generated recommendations.

The path to autonomy will be incremental: companies will first augment human decision-making, then automate routine and semi-structured decisions as governance, trust, and data maturity improve.

The next frontier is the Autonomous Enterprise

The Autonomous Enterprise is an operating model where AI workflows, contextual business data, and embedded governance work together to anticipate disruption, coordinate action, and continuously improve performance.

The shift is moving from isolated copilots to coordinated agent-to-agent workflows spanning the supply chain.

In autonomous production environments, supplier reliability agents can monitor vendor risk while workforce orchestration agents align labor capacity with demand. Procurement agents execute sourcing decisions, and production planning agents dynamically rebalance schedules in response to changing conditions.

A similar pattern is emerging in asset management, where alert-processing, maintenance, warehouse replenishment, and goods-movement agents collaborate to resolve operational issues with minimal human intervention.

The business impact is significant. Agentic AI has by 20 to 30%, , and helped .

Collectively, these improvements mark the transition from reactive supply chains to systems that can increasingly anticipate, decide, and execute autonomously.

Building the autonomous supply chain

Capturing this opportunity requires three capabilities that remain fragmented in many organizations today:

• Organizational intelligence: The ability to detect patterns, anticipate risks, and reason across constraints
• Contextual data: Trusted operational data, business rules, workflows, and policies that ground AI decisions in enterprise reality
• Embedded execution: Integrating intelligence directly into workflows so actions can move from recommendation to execution without manual intervention

This creates a virtuous cycle: better data improves decisions, better decisions improve processes, and improved processes generate richer operational data over time.

Importantly, companies do not need to rebuild the enterprise from scratch. Deterministic systems of record remain essential for control, compliance, and auditability. The real transformation lies in rewiring how decisions are made and governed.

Organizations moving fastest are focusing first on high-value, high-frequency decisions such as forecasting, inventory optimization, disruption sensing, transport planning, procurement workflows, maintenance, and customer-service resolution.

The bottom line

The future of supply chain management will not be defined by more digital tools alone. It will be defined by the ability to operate the supply chain as a connected, adaptive, and increasingly autonomous system.

For leaders who move first, supply chain will evolve from a cost-management function into a competitive differentiator, enabling faster time to market, stronger service levels, and greater resilience. The organizations that lead will not be those running the most AI pilots. They will be the ones using AI to redesign how the enterprise senses, decides, and acts across the end-to-end supply chain.

For more information about Autonomous Supply Chain Management, download the white paper, .

Hagen Heubach is chief marketing officer for Supply Chain Management at SAP.

While expectations for reliable, on-time delivery remain high, organizations are navigating faster‑changing demand, more complex global networks, and increasing pressure on cost and working capital. And they’re looking for ways to turn insight into action more quickly and consistently across the supply chain.

51·çÁ÷has been helping organizations build more connected and intelligent supply chains for over 50 years. At 51·çÁ÷Connect in October, we introduced 51·çÁ÷Supply Chain Orchestration, establishing a foundation for detecting issues, coordinating responses, and connecting execution across complex supply networks.

The innovations announced this week at 51·çÁ÷Sapphire extend that vision further. By introducing a new set of AI-driven assistants and agents, we’re moving orchestration toward an autonomous operating model, where planning, manufacturing, logistics, and asset operations increasingly anticipate, coordinate, and resolve without manual intervention at every step.

AI grounded in real operations

AI delivers lasting value in supply chain management only when it is embedded where work actually happens. Autonomous agents do not operate independently of enterprise applications; they rely on deeply integrated processes and trusted data. Precision, compliance, and resilience depend on this foundation. Without it, AI does not scale or earn trust.

At SAP, the Autonomous Enterprise represents a vision for how organizations will run their businesses in the future: with insight, decision-making, and execution increasingly connected, while people remain firmly in control. Autonomous Supply Chain Management is a practical step toward that vision.

Autonomous Supply Chain Management reflects an evolution in how planning, execution, and operations work together. People define goals and priorities, assistants orchestrate activity across domains, and agents execute the work—all within governed, end‑to‑end processes.

At 51·çÁ÷Sapphire, we’re introducing , enabled by new Joule Assistants and Industry AI scenarios that apply this model to daily operations across planning, manufacturing, logistics, engineering, and asset management. General availability will be phased throughout 2026, starting now.

Joule Assistants across the supply chain

Rather than disconnected AI tools, the following assistants will be embedded directly into core 51·çÁ÷supply chain applications, where deep process knowledge, semantically rich business data, and enterprise‑grade governance already exist.

Each will support a distinct area of responsibility while sharing context, data, and outcomes across the supply chain:

• Asset and Service Assistant: Changes how work gets detected and dispatched, turning signals and anomalies into action rather than queue items
• Business Network Assistant: Extends this coordination outward across suppliers, logistics providers, and service partners so execution doesn’t stall at the edges of the enterprise
• Logistics Assistant: Keeps warehouse and transportation execution moving as conditions change, coordinating agents rather than waiting for human handoffs at every step
• Manufacturing Assistant: Connects shop floor signals with broader operational context so teams can act on disruptions faster
• Planning Assistant: Helps planners stay ahead of exceptions and constraints without having to manually piece together signals from across the network
• Product Design Assistant: Helps engineering and manufacturing teams stay aligned as products evolve, surfacing the downstream implications of changes before they create rework or delays

From assistants to autonomous agents

In addition to these assistants, 51·çÁ÷is delivering more than 60 purpose‑built agents across supply chain processes. These agents are designed to sense events, analyze impact, and take guided action within defined business guardrails, helping coordinate execution while keeping people firmly in control.

In manufacturing, agents such as the Production Excellence Agent and Production Master Data Readiness Agent continuously monitor production, quality, and machine signals to detect issues early and keep routings and work instructions aligned with enterprise plans. In asset and service operations, the Asset Performance Alert Processing Agent and Technician Briefing Agent are designed to assess asset conditions, prioritize work, and increase first time fix rates, helping reduce downtime and improve responsiveness.

Beyond supply chain-specific scenarios, these assistants and agents will also extend into SAP’s cloud ERP environment, including , supporting SAP’s broader Autonomous Enterprise strategy. General availability will be phased through 2026, starting now.

Building on this foundation, 51·çÁ÷Industry AI adds industry-specific intelligence that complements the core assistants. Rather than standalone features, Industry AI brings together purpose-built agents, process expertise, and business data to drive measurable outcomes. This value-led approach helps organizations apply AI in ways that reflect regulated requirements, complex production models, and asset-intensive operations – accelerating information across entire industry value chains.

People remain responsible for strategy, oversight, and the decisions that require judgement. What changes is how consistently high-volume, time-sensitive coordination happens across the supply chain.

Where this shows up in practice

The Autonomous Enterprise is our vision, and the innovations we’ve announced at 51·çÁ÷Sapphire are concrete steps that customers can build on within current 51·çÁ÷environments. They are focused on addressing value leakage caused by fragmented handoffs, delayed decisions, and manual work.

In planning, new capabilities will connect commercial decisions directly with supply planning, linking promotion and pricing plans to inventory and replenishment to reduce stockouts, minimize write-offs, and improve planning consistently. New capabilities include vendor-managed inventory, transportation load building, deployment optimization, and co- and by-product planning.

In manufacturing and engineering, updates to will strengthen compliance and traceability in regulated environments. AI capabilities in the engineering-to-manufacturing handover will help teams understand the downstream impact of design changes before they reach the shop floor, surfacing implications for bills of materials, routings, lead times, and costs directly in context.

In , new Joule Agents will support execution-level decisions across warehouse and transportation operations, validating inbound receipts, aligning labor with real workload, and helping organizations respond faster to shifting constraints. Predictive labor planning in will allow operations teams to anticipate workforce needs rather than react to gaps.

In asset and service management, a new 51·çÁ÷Field Service and Asset Management solution will bring planning, scheduling, dispatching, and field execution together in a single experience, connected to so work execution, parts usage, and costs stay aligned across service, operations, and finance.

These capabilities will become available in phases through 2026, aligning with customers’ existing 51·çÁ÷landscapes. Together, they represent incremental but meaningful progress toward more connected, automated, and resilient supply chain operations.

The path forward

Supply chains don’t become autonomous overnight. This evolution happens workflow by workflow, expanding automation where it delivers real value, while keeping people firmly in control. As AI becomes embedded in execution, supply chain teams spend less time monitoring and firefighting, and more time shaping decisions, managing trade-offs, and building resilience.

This shift is bigger than any single organization. In a new white paper, , we explore how leading organizations are moving beyond isolated AI pilots toward AI embedded across end-to-end supply chain processes, and what it takes to get there. This article draws on multiple sources, including analytical support from McKinsey & Company.

That’s the direction we are moving, from reacting toward supply chains that anticipate, absorb, and adapt. What we’re introducing at 51·çÁ÷Sapphire reflects that commitment.For more details on all announcements made this week, please refer to the .

Dominik Metzger is president and chief product officer of SAP Supply Chain Management.

In his kickoff keynote at 51·çÁ÷Sapphire Orlando, Florida, Klein and other 51·çÁ÷Board members detailed how 51·çÁ÷plans to bring agentic AI to the world’s most critical business workflows so that humans and AI can meet the accelerating demands of global business profitably, strategically, and safely.

“Today I’m super proud to launch our new 51·çÁ÷Business AI Platform, which forms the basis for our vision of the future of business: the Autonomous Enterprise, where agents run the business and you can focus on what truly matters,” Klein said.

Enterprise AI is at an inflection point, Klein told his 30,000-strong in-person and virtual keynote audience, and 51·çÁ÷is in a unique position to deliver what customers need to turn their businesses into autonomous enterprises.

Click the button below to load the content from YouTube.

Always allow YouTube

The business AI imperative

Across industries, organizations are investing heavily in artificial intelligence, yet many still struggle to translate that investment into meaningful business value. At 51·çÁ÷Sapphire, the message was clear: This isn’t a technology problem; it’s a context and execution problem.

While 80% accuracy may be sufficient for consumer AI applications, Klein said, “Eighty percent is just not good enough when you run the world’s most business-critical businesses. They [LLMs] should not guess; they should deliver accurate, compliant, and secure outcomes.” 

Klein acknowledged that while adoption of AI has become near-universal, tangible business value remains elusive. Citing a recent Stanford AI survey, he noted that almost every company is now using AI, but seeing only limited return.

The reason, he argued, lies in a structural gap. Above the waterline of enterprise AI, LLMs continue to improve at tasks trained on publicly available data, while below it lies what enterprises truly need: AI that understands mission-critical business data, end-to-end processes, and operates within security, compliance, and governance frameworks.

ERP as the foundation for business AI

SAP’s answer to this challenge begins with what Klein described as “the brain of every company: its ERP system.” For over 50 years, 51·çÁ÷has had solutions with incredibly deep process and data domain know-how alongside the governance requirements, compliance controls, and company-specific configurations that define how businesses actually run.

Now, as part of the company’s new vision, 51·çÁ÷plans to infuse this institutional knowledge into AI agents, enabling them to navigate thousands of business processes, select from more than 7 million data fields, and verify identity and access authorizations before returning any output.

“We’re bringing together LLMs with 50 years of business know-how stored in our ERP. But to do this, we had to do nothing less than completely reinvent our company,” he told the audience. “Today we are very excited to show you the new 51·çÁ÷and our vision for the Autonomous Enterprise.”

51·çÁ÷Business AI Platform

To bring this vision to life, 51·çÁ÷executives on stage announced a series of important innovations, beginning with the launch of the new 51·çÁ÷Business AI Platform, a unified architecture bringing together 51·çÁ÷Business Technology Platform, 51·çÁ÷Business Data Cloud, and AI Foundation under a single roof.

“The heart of this new platform is the rich context layer,” said Klein. “Here, we infuse the deep ERP business domain know-how into the AI agents. Through our knowledge graphs, our AI agents have now a compass, a map, to find the right process and data in your ERP universe. And to provide the agents even more context, we are also introducing our new 51·çÁ÷Domain Models. They have been trained on SAP’s code to even better understand the business logic of your company.”

But, he said, 51·çÁ÷is going further: “Because you run your business not only with 51·çÁ÷solutions, our AI agents have to also understand non-51·çÁ÷data. That’s why we included our 51·çÁ÷Business Data Cloud in the context layer to build a single semantical data layer across 51·çÁ÷and non-SAP. No more silos, no spaghetti data sprawl—because no AI agent can compensate for a broken data model.”

Echoing Klein, 51·çÁ÷CTO Philipp Herzig, who presented the platform in detail, said it has been designed to close the agent adoption gap in the enterprise by delivering outcome, speed, enterprise-readiness, and context. “It’s the place where you build, contextualize, reason, and govern AI,” he said.

Herzig explained that the platform is structured around three layers: the context layer which Klein referenced, the build layer, and the governance layer. “Agents are only as powerful as the context they operate on,” he said. “Lacking context is the number one reason why enterprise AI projects fail to deliver value.”

Within the build layer of the new platform, the new Joule Studio is designed to understand a company’s business challenges and enables the building of new AI agents quickly and easily.

The third tier is the governance layer, anchored by the new 51·çÁ÷AI Agent Hub built on 51·çÁ÷LeanIX. This provides a single command center to discover, manage, and govern all AI agents—51·çÁ÷and non-SAP. It will be generally available in Q3 and included in 51·çÁ÷Business AI Platform at no additional charge.

Underscoring the changing AI marketplace, Herzig was joined on stage by KPMG Global Head of Advisory Rob Fisher, who told the audience: “What I’m hearing from clients is a clear shift; they’re moving from AI pilots to embedding integrated AI and agents into how work gets done. Where we see leaders really separating from the pack is in the execution and the organizational adaptability.”

51·çÁ÷Autonomous Suite

Building on the platform, 51·çÁ÷Executive Board Member Muhammad Alam, 51·çÁ÷Product & Engineering, announced the transformation of SAP’s SaaS application portfolio into the 51·çÁ÷Autonomous Suite, described as the most significant evolution of SAP’s applications business in the company’s history.

The suite spans five domains: Autonomous Finance, Autonomous Spend, Autonomous Supply Chain Management, Autonomous HCM, and Autonomous CX, with more than 200 agents and over 50 assistants available in the coming months. Each assistant is mapped to core business roles and carries defined KPIs tracked through 51·çÁ÷AI Agent Hub.

“51·çÁ÷Autonomous Suite brings together the depth of our process expertise, semantically rich data, and built-in governance and compliance,” said Alam. “These agents are designed with outcomes as a core objective. Each assistant has a defined set of ROI KPIs that you can expect it to deliver.” 

“Underpinning the autonomous suite are out-of-the-box agents—hundreds of agents cutting across all core business processes,” he shared. “These agents come together into what we call assistants, or Joule Assistants. We’ve mapped these assistants to roles across the core processes of an organization, because we know that the first step 
in realizing value from AI is to empower your people to do more, do it better, or do things that just weren’t possible to be done before.”

Turning to Joule itself, Muhammad said 51·çÁ÷is fundamentally reimagining how users will interact with 51·çÁ÷applications in the future.

“We call this Joule spaces and along with the familiar Joule conversations experience and Joule Studio 2.0, it is now part of what we call Joule Work,” he explained.

“Joule Work represents a massive step forward in super-charging the capabilities of Joule as we know it today,” Alam said. â€śWith Joule Work, we’re bringing a claw-based agentic harness to Joule along with computer and file access, better support for open standards such as MCP and A2A, access to a more complete knowledge base, and, of course, amazing visualizations on the fly.”

Industry AI: H&M and Sector-Specific Transformation

During the keynote, 51·çÁ÷Chief Operating Officer Sebastian Steinhaeuser introduced the Industry AI initiative, delivering AI-powered solutions built on decades of sector-specific expertise across 26 industries. In life sciences, he highlighted how 51·çÁ÷customer Takeda is achieving up to 10% productivity gains, up to 25% reduction in revenue loss from stock-outs, and up to five percent reduction in safety stock through Autonomous Regulated Manufacturing.

He was also joined on stage by H&M Group CDIO Ellen Svanström, who discussed how the fashion retailer is embedding AI across its value chain. Built on RISE with SAP, 51·çÁ÷Business Data Cloud, 51·çÁ÷Commerce Cloud, and 51·çÁ÷SuccessFactors solutions, H&M has developed a Store Intelligence Agent that processes real-time signals to generate actionable recommendations for store managers. Svanstrom also demonstrated the AI-powered InStore Concierge, a customer-facing agent that bridges digital and physical retail through personalized outfit recommendations and real-time availability.

RISE with 51·çÁ÷and 51·çÁ÷GROW: Path to the Autonomous Enterprise

Returning to the keynote stage, Klein emphasized that technology adoption alone does not create business value. Simply plugging AI agents into your system landscape will drive zero value, he said. “Moving to the Autonomous Enterprise requires serious change management. Adoption of AI goes hand-in-hand with business process change and end user enablement.”

To support customers on this journey, 51·çÁ÷announced a comprehensive reset of its RISE with 51·çÁ÷and 51·çÁ÷GROW offerings. RISE with 51·çÁ÷customers will receive contractual commitment to activate three Joule Assistants within the first year, with the Max Success Plan extending adoption across the full enterprise.  

51·çÁ÷GROW customers will receive more than 20 AI assistants from day one, with an AI-enabled toolchain designed to support go-live in weeks. New partnerships with Palantir and Accenture will support the most complex migration scenarios.

Closing: The Autonomous Enterprise

Klein closed the keynote by asking Joule to summarize the key takeaways and noting that 51·çÁ÷is evolving from being a software company to becoming a business AI company.

“We showed how to turn the promise of business AI into reality with 51·çÁ÷Business AI Platform, which provides the data processes and governance AI need to deliver accurate and secure outcomes at scale; we introduced the 51·çÁ÷Autonomous Suite, where applications reason, decide, and act for you; and we showed how to manage change management with RISE with SAP. Together with customers and partners, we showed how 51·çÁ÷is helping companies realize the vision of the Autonomous Enterprise.”

“We’ve been reinventing how businesses run for over 50 years, and now by infusing SAP’s ERP brain into the new 51·çÁ÷Business AI Platform, we’re solving one of the biggest challenges businesses are facing today: how to turn AI into business value,” he said. “It’s the end of long negotiations, supply chain disruptions, financial blind spots, and the beginning of better: Welcome to the Autonomous Enterprise.”