📊 Full opportunity report: Why SAP Is Betting On In-House AI Systems Instead Of Outsourced Brainpower on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
SAP is increasingly developing its own AI systems, such as Joule, to own and control enterprise data, rather than outsourcing AI model development. This shift aims to strengthen its position in enterprise AI and reduce reliance on external providers, but faces adoption and cost challenges.
SAP is intensifying its focus on in-house AI systems, with the launch and expansion of its Joule platform, designed to embed AI deeply into its enterprise software. This strategic pivot underscores SAP’s commitment to owning the data and infrastructure that underpin AI capabilities, rather than relying on externally sourced models. The move is a response to the evolving landscape of enterprise AI, where control over data and context is increasingly critical.
As of mid-2026, SAP’s Joule AI layer is integrated across more than 35 solutions, including S/4HANA Cloud, SuccessFactors, and Ariba, with over 30 specialized agents and 2,500+ ‘Joule Skills.’ SAP has committed €100 million to a partner fund aimed at enabling systems integrators to develop custom AI agents using Joule Studio, its low-code agent builder. The company claims measurable improvements: a global retailer reduced HR process cycle times by 40–60%, an Argentine airport operator cut direct costs by 16% and administrative effort by 90%, and developers reported 20% productivity gains.
SAP’s architecture leverages a Knowledge Graph that reads business metadata directly from its Business Technology Platform, allowing Joule to understand context-specific workflows—an advantage over frontier models that pull answers from open internet sources. This design makes SAP’s AI model-agnostic, consuming third-party models and orchestrating them within its own platform, thus maintaining control over the AI ecosystem.
Own the system of record.
Rent nobody’s brain.
SAP’s AI bet is the incumbent’s inversion of the frontier race: don’t build the smartest model — own the data smart models are useless without, and meter access through Joule, an orchestration layer indifferent to which model wins.
The stack — where SAP chose to stand
You can switch AI vendors in an afternoon. You cannot switch your general ledger.
Honest bull / bear
Bull
- Best data-layer position of any incumbent — the one place hyperscalers can’t reach
- Knowledge Graph is context no model scale substitutes for
- Model-agnostic: owns the layer above commoditizing models
- Named, operational customer outcomes (40–60% HR cycle time, 90% admin cut)
Bear
- Consumption pricing is hard for CFOs to forecast — adoption stalls
- “Activated” ≠ “adopted”: the €100M fund admits demand needs subsidizing
- Depends on frontier models it doesn’t control
- Innovation tax: everything must work across a regulated installed base
Mastering Enterprise Platform Engineering: A practical guide to platform engineering and generative AI for high-performance software delivery
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Implications of SAP’s In-House AI Strategy for Enterprise Software
SAP’s shift to develop and deploy its own AI systems positions it uniquely in the enterprise AI market. By owning the data layer and integrating AI directly into its core solutions, SAP aims to create a more trustworthy, context-aware, and controllable AI environment. This approach could give SAP a competitive advantage over hyperscalers and frontier labs, which primarily focus on model development rather than data ownership. However, it also introduces risks related to adoption, cost management, and dependence on third-party models, which could impact the effectiveness and scalability of SAP’s AI initiatives.
SAP’s Enterprise AI Evolution and Strategic Positioning
Throughout 2025 and into 2026, SAP has emphasized integrating AI into its existing enterprise solutions, aiming to embed AI agents as first-class users alongside humans. The company’s strategy contrasts with the broader industry trend of building larger, more generalized models; instead, SAP focuses on owning and utilizing its structured, permissioned enterprise data. The launch of Joule and the €100 million partner fund reflect SAP’s intent to accelerate this approach, building a robust AI ecosystem tightly coupled with its core business processes.
Previously, SAP’s AI efforts relied heavily on integrating external models, but recent developments highlight a shift toward internal development and orchestration, leveraging the Knowledge Graph and third-party foundation models like Prior Labs’ offerings. This evolution is driven by the need for trustworthy, compliant AI that can operate reliably within mission-critical enterprise environments.
“Our goal is to own the data that models need, not just build the smartest model. Control over the data layer is our strategic advantage.”
— SAP Executive
Uncertainties Around Adoption and Cost Management
It remains unclear how quickly and broadly SAP’s customers will adopt Joule at scale, given the current challenges of variable AI consumption costs and the need for organizational change. The effectiveness of SAP’s strategy depends on customer willingness to reduce custom code and migrate to standard data structures, which may face resistance. Additionally, reliance on third-party models introduces dependency risks if model quality or access conditions change unexpectedly.
Next Steps in SAP’s AI Ecosystem Expansion
SAP plans to expand Joule’s capabilities, aiming for 50 assistants and 200 agents by Q3 2026, supported by its partner fund and new developer tools. The company will likely focus on increasing customer adoption, refining cost models, and demonstrating ROI through case studies. Monitoring how organizations operationalize Joule and manage AI costs will be critical to assessing the long-term success of SAP’s in-house AI approach.
Key Questions
Why is SAP focusing on in-house AI systems instead of outsourcing?
SAP aims to own its enterprise data and infrastructure, ensuring greater control, trustworthiness, and context-specific AI capabilities that external models cannot easily provide.
What are the main risks associated with SAP’s AI strategy?
The key risks include unpredictable AI consumption costs, dependence on third-party models, slow customer adoption, and challenges in migrating to standardized data structures.
How does SAP’s Knowledge Graph enhance its AI capabilities?
The Knowledge Graph allows Joule to understand business-specific workflows and legal contexts, making AI responses more accurate and relevant to enterprise needs.
Will SAP’s approach be disruptive to the enterprise AI market?
By owning the data layer and integrating AI deeply into its core solutions, SAP could reshape enterprise AI, emphasizing control and trust over model size and performance.
What’s next for SAP’s AI development in 2026?
SAP plans to expand Joule’s capabilities, increase customer adoption, and demonstrate measurable ROI, while managing costs and dependency risks.
Source: ThorstenMeyerAI.com
