The SAP advantage for generative AI

Driving gen AI success with SAP systems

Organizations seek generative AI initiatives that can learn, adapt, and revolutionize operations, products, and services at an unprecedented scale. Enter SAP systems, an ideal resource because:

• They dig deep. SAP systems yield a high volume and variety of business data.
• They’re well organized. The inherent structure of SAP systems empowers gen AI models to process information more efficiently and extract valuable insights.
• They’re trusted. SAP data is accurate, well maintained, and secure.
• They’re connected. Well-architected SAP platforms enable data sharing.

New research from the IBM Institute for Business Value (IBM IBV) and the SAP Research Insights Center (SAP RIC) shows that SAP data is a gold mine for generative AI initiatives. We conducted three surveys of more than 1,200 executives over a seven-month span to understand how their organizations are implementing gen AI solutions across their SAP platforms and data.

Our results reveal a distinctive advantage for those organizations with the most mature capabilities in SAP-specific generative AI. They experience a boost in profitability, outpacing their peers. Organizations integrating gen AI strategically across business areas—versus using a piecemeal approach—are particularly far ahead.

Organizations with the most mature capabilities in SAP-specific generative AI outpace peers in profitability.

Executives increasingly recognize the value in linking SAP systems to AI. More than half of generative AI spend is being directed to SAP-specific initiatives, our research shows. In fact, 56% of executives say they plan to divert spending from other budget areas to SAP-specific gen AI, with one-third saying that SAP systems hold the most data critical to their success. These encouraging findings suggest AI innovation leveraging SAP systems will bear fruit for those who adopt and adapt most quickly.

What follows is a roadmap on how best to leverage SAP-specific gen AI initiatives. In part one we explore key prerequisites, from cloud infrastructure to skills to governance. In part two we delve into strategic investments, including specific business areas that are experiencing early success. Both part one and part two are punctuated by examples that bring SAP-specific generative AI to life. Finally, we offer an action guide that spans tech, talent, and strategy to help drive the most value from SAP-specific gen AI implementation.

Case study

Essential ingredients for gen AI success

The advantage of cloud infrastructure

Gen AI demands an architecture that can handle vast volumes of complex data while delivering significant, scalable computing power and advanced security. Cloud environments are a natural complement, serving as a foundation for a security-rich, centralized platform for data management and collaboration. A hybrid cloud SAP architecture provides access to expensive GPUs only when needed, as well as the flexibility and scalability to switch between models and services.

While organizations typically deploy their SAP systems in multiple environments, many are increasingly opting for a cloud or hybrid deployment to gain scalability and integration capabilities. Executives report their SAP systems are either hosted in a hybrid environment comprising cloud and on-premises resources (41%) or are fully hosted in the cloud (42%).

Organizations leveraging cloud for SAP systems report positive results: 44% say the architecture has helped them move faster with their gen AI initiatives versus 12% who say it is slowing them down; the rest report no impact.
 

Cloud architectures are helping accelerate SAP-specific generative AI adoption.

Nearly two-thirds (60%) say their SAP infrastructure is now optimized to support their SAP-specific gen AI goals. Those organizations with the most advanced SAP cloud architectures—including RISE with SAP and SAP S/4HANA—are experiencing 10% higher revenue growth than those with less mature architectures.

For those not fully on cloud, the urgency is real. Many organizations (67%) running SAP are accelerating that process specifically to capture the benefits of SAP-specific gen AI. This is up from 59% in March 2024, suggesting that executives see the need for gen AI-ready cloud infrastructure to unlock SAP system potential.
 

Perspective

Identifying skills, building talent 

For generative AI initiatives to succeed, having the right talent can be as important as the technology itself. While gen AI can amplify human capabilities, it often requires refreshed skillsets.

Previous research from the IBM IBV found that executives increasingly value soft skills over technical skills. Data literacy and machine learning fundamentals such as model training and prompt engineering are, of course, core capabilities. But executives rank problem solving/crisis resolution as the skill affected the most by the emergence of generative AI. Business communication and agility are next on the list.

In our new survey, 69% of executives say their organization has the right skills to extract value quickly from their SAP-specific generative AI initiatives, up 14 percentage points from four months earlier. Similarly, 58% consider their AI talent and skills to be exceptional or best-in-class. Among those that have successfully embedded or integrated SAP-specific gen AI across multiple business functions, 92% say they have the necessary gen AI talent and skills.
 

Organizations that strategically apply SAP-specific gen AI are more confident in their workforce skillsets.

Organizations of all types can fill the AI skills gaps through focused reskilling, hiring, and working with business partners. What’s more, gen AI itself can help ramp up skills by accelerating adaptive, personalized learning. For example, it is already revolutionizing corporate finance training by enabling personalized learning experiences through tutors powered by large language models (LLMs) that adapt to each learner’s pace, style, and comprehension level. And this approach can be applied to virtually any business area, including technology.

Organizations should also focus on alleviating employees’ concerns about being replaced by AI, emphasizing that AI is intended to help them work faster and smarter, not take their jobs. Offering employees ways to experiment with SAP data and gen AI tools—with proper security precautions and governance in place—can spark creativity and confidence as well as greater buy-in and ultimately greater success.
 

Perspective

Getting ahead on governance

Gen AI introduces an entirely new realm of ethical and trust considerations—which means organizations need updated governance, risk, and compliance measures. Leaders can’t focus on only what gen AI can do. They must talk about what it should do. Is sensitive data protected? Does training data respect copyrights? Are outputs biased or just wrong?

Generative AI governance challenges transcend the entire enterprise. But even high-quality ERP data may have hidden biases—or the AI model itself can introduce biases based on how it’s trained. Organizations need a solid understanding of how the model arrives at the content it creates. Meanwhile, the data and the models must be protected from emerging cyber threats that can alter both inputs and outputs for malicious intent.

These challenges may be hindering adoption: the percentage of executives who said they will forego the benefits of SAP-specific gen AI because of ethics concerns surged from 26% in March 2024 to 65% in July 2024. This pivot suggests an urgent need for tighter gen AI governance and explains widespread adoption by SAP and others of UNESCO’s 10 guiding principles for ethics in AI.
 

Executives recognize that their AI governance capabilities are not up to the emerging generative AI challenge.

To increase enterprise awareness of ethical and regulatory issues and establish actions for how to address their impacts, organizations need to establish a comprehensive gen AI governance framework. Emerging AI regulations such as the EU AI Act offer more clarity around degrees of risk and corresponding responsible AI requirements. Enterprises that already face dynamic regulatory environments are better positioned to understand the task of SAP-specific gen AI governance.

AI governance best practices must extend beyond simply complying with regulations to encompassing a more robust system for monitoring and managing AI applications. For enterprises using SAP systems, that can mean updating functional workflows to account for the new guidance and building in flexibility to adapt as regulations unfold.

AI governance should also enable broad oversight and control over AI systems. Examples of elements to include are visual dashboards that provide real-time updates on the health and status of AI systems, automated monitoring and performance alerts for model bias and drift, and audit trails.

Most executives (70%) say their organizations have clear policies about the proper use of gen AI. That includes not allowing SAP data to be used in publicly available gen AI tools. These are positive first steps, but policies and direction are not as strong as formal governance structures. Organizations must build an internal culture focused on responsible AI, embracing employee education, championing ethics teams and policies, and eagerly responding as regulatory requirements evolve.

Setting gen AI in motion

Invest in the sweet spots

Organizations are betting on gen AI to unleash the immense, transformational potential of their SAP systems. They are considering a treasure trove of choices for SAP-specific generative AI use cases across business areas—from customer service to finance to HR to marketing—and are narrowing the list based on feasibility and value.
 

Organizations are evaluating SAP gen AI use cases based on feasibility and value.

Executives say gen AI use cases with the highest value fall in areas with the largest employee populations, such as customer service (53%), supply chain (51%), and sales and marketing (47%). Similarly, they consider use cases in these same business functions to be the most feasible.

Gen AI can have significant impact on productivity in these areas. For example, an AI-powered assistant for automotive warranty claims can accelerate the warranty process for both consumers and dealers by pre-filling forms and scheduling the inspections at the nearest dealership. Likewise, embedding gen AI into supply chain planning tools can equip employees with a personal assistant that answers queries by pulling data from the SAP platform and other sources, expediting decision-making. In sales and marketing, gen AI can streamline creation of personalized communications and other content.

Among these high-value, high-feasibility areas, organizations report making the most progress in marketing, where 18% are either operationalizing or optimizing their AI use cases. A slightly smaller cohort reports gains in research and innovation (16%) and information technology (16%), where gen AI could be helping employees convert data to designs more quickly or accelerating software development.

Over the next three years, executives say they plan to expand AI use cases into physical domains such as manufacturing. For instance, industrial manufacturers can combine their vast data and SAP digital manufacturing analytics with machine and equipment modeling in an AI-augmented data retrieval approach to detect maintenance issues and retrieve resolution recommendations faster. Organizations could also optimize production and inventory or generate demand forecasting.
 

Perspective

From bet to boom

While gen AI solutions are still in the early days of implementation, the use of SAP systems is already boosting operating margins. We developed an index that assesses the strength of organizations’ SAP-specific gen AI business and technical capabilities, including their AI strategy and vision, data and technology, talent and skills, engineering and operations, operating model, and organization.
 

More advanced SAP-specific generative AI capabilities translate to greater profitability.

We found the most mature organizations experience an average of 20% profit margins compared to an average of 16% for their peers. We project that pattern will continue through 2026 with an average of 23% profit margin for more mature organizations versus an average of 19% for the rest. (We observed a similar difference in profit margins in the pre-generative-AI days, which suggests that early AI gains have been used to fuel generative AI investments.)

The results on ROI are more nuanced. Executives report 6.8% ROI on SAP-specific gen AI initiatives. Our data shows an average of 48% of organizations are testing use cases across 11 business areas. Given the cost and complexity of gen AI solutions, organizations should refine their investments and take a more strategic approach that integrates generative AI into their SAP platform across functional areas rather than limiting it to small, siloed areas. Organizations project making gen AI an integral part of their operations and expect to nearly double their ROI to 12.2% from generative AI investments in 2025.

While enterprises may be tempted to try to customize each facet of their generative AI solutions, they also understand the inherent advantages associated with a more homogeneous approach. SAP and IBM have deeply embedded AI and generative AI into business processes, enterprise analytics, and development of future AI-enabled solutions across the SAP platform. Collectively these solutions enable SAP users to interact with enterprise systems in more natural ways, asking questions of and interacting with enterprise data to improve their productivity. The generative AI Hub in SAP AI Core allows AI solutions to be developed, deployed, and managed in a standardized, scalable way with full lifecycle management, allowing organizations to confidently make data-driven decisions.

Action guide

SAP systems are an incredibly valuable tool for bringing generative AI initiatives to life. Two-thirds of executives say data processed by and contained in their SAP systems should serve as the starting point for most gen AI endeavors, and 56% report their organization began their gen AI journey with SAP data. How do you optimize gen AI investments and outcomes? By putting the proper base in place and strategically investing in an integrated approach. Here are three priorities—focused on data, talent, and strategy—that can put those goals into action.
 

Plan like a seasoned Chief Data Officer (CDO)

Embrace that SAP-specific gen AI success demands trusted data at scale across all departments and SAP functional areas.

• Audit your SAP data landscape. Identify the data available, its quality, and accessibility.
• Evaluate whether your existing infrastructure can support gen AI initiatives. Intentionally design a cloud architecture that supports gen AI innovation.
• Examine your cybersecurity capabilities to see if they are strong enough to protect your gen AI applications, data, and models.
• Integrate and harmonize SAP data across enterprise silos to ensure that SAP-specific gen AI solutions are enabled by the real sources of enterprise truth.

Prioritize like an effective Chief Executive Officer (CEO)

Accept that new and emerging SAP-specific gen AI tools will enable new, and as yet, unknown enterprise solutions.

• Identify business problems that can be solved with gen AI solutions today and into the unknown future. Clearly define expected outcomes in each area over time.
• Align available SAP data to the business problems and define the most feasible and valuable use cases.
• Encourage and reward SAP-specific gen AI experimentation to increase enterprise learning and drive near-term wins.
• Empower functional owners, process teams, and individual employees to use SAP-specific gen AI outputs as inputs to business outcomes they will continue to own.

Personalize like an astute Chief Human Resources Officer (CHRO)

Embody in your organization the notion that only trust in the enterprise—its mission, its data, its people—will enable future success.

• Establish or update your AI governance framework. Outline ethical considerations, data privacy practices, and accountability mechanisms.
• Offer training or workshops to all employees on basic AI concepts, data literacy, and responsible AI practices.
• Map gen AI skillsets to roles. Share the roles with employees and develop training plans to help them develop in their current jobs or move into new positions.
• Foster an experiment-friendly culture that equally celebrates success—and failures that lead to future success. Encourage employees to experiment with SAP-specific gen AI within the well-defined boundaries of good governance and trusted data.