Who Needs AI With ERP?

Resilience, observability, and governance

Many AI experts in the SAP community believe that resilience will become a key indicator of digital AI performance. Large language models (LLMs), such as ChatGPT from OpenAI and Gemini from Google, are based on highly sophisticated statistical methods and machine learning. However, deterministic algorithms, such as those used in SAP's ERP software for over 50 years, are not intended for LLMs. Therefore, there is always a risk of hallucinations when using traditional AI.

Due to minor disruptions, AI systems can quickly escalate across applications, cloud regions, payment systems, and external services. However, ERP users demand reliability, availability, and security—in other words, resilience and observability. These requirements can no longer be viewed as separate disciplines but rather as a common AI goal. SAP has yet to provide proof that an S/4 BTP system can absorb disruptions, react quickly, and ensure a consistent user experience under load.

The AI community problem

A study commissioned by Dynatrace and FreedomPay reveals the fragility of digital ecosystems and the rapid escalation of customer frustration and financial losses from technical failures. In the UK, an estimated 1.6 billion GBP in revenue is at risk from payment failures each year, and around 1.9 billion EUR is at risk in France. A single disruption can spread across networked ERP and AI systems, highlighting how tightly coupled modern end-to-end processes are.

Customers immediately feel the effects of these failures. Their patience wears thin within a few minutes, and many cancel ERP transactions if the problem persists for more than fifteen minutes. However, the average outage lasts more than an hour—by which time the damage is already done. Almost a third of customers lose trust after just one incident, particularly younger, digitally savvy target groups.

According to Dynatrace, this situation requires a common approach to resilience. Companies need a shared understanding of how ERP services behave, how errors propagate, and how recovery affects the customer journey. Resilience is measured by how AI and ERP systems respond under pressure, not just how they perform during normal operations.

2026: CES (Las Vegas, USA) and Viva Technology (Paris, France)

Events, congresses, and trade fairs are an important part of the AI discourse. During the second week of January, CES 2026—the world's most powerful tech event—took place in Las Vegas, where leading IT industry managers and their customers met. Last year, Viva Technology, the European AI hotspot, was in Paris in June. Of course, Nvidia was well represented there. However, former SAP CEO and current Siemens Supervisory Board Chairman Jim Hagemann Snabe also appeared on the keynote stage. Although SAP was absent from Viva Technology in Paris in 2025, almost all of SAP's major consulting partners were represented.

Germany is the preferred host country for Viva Technology 2026. Whether SAP will find its way to Paris this year remains to be seen. However, if SAP CEO Christian Klein is serious about AI, he should follow Snabe's lead and appear on the Viva Technology keynote stage.

The AI challenge of agentic AI and observability

No IT company, whether established or a startup, will likely be able to overcome AI challenges alone. SAP will also need public discourse on AI and IT partners. SAP's solo efforts with partial partnerships have not led to overwhelming AI successes. The subject matter is difficult to master!

Agentic AI exemplifies the power of AI, but it is much more difficult to master than traditional LLMs and machine learning systems. Even a well-structured digital environment can tip over into unpredictable behavior when AI agents on the SAP BTP (Business Technology Platform) coordinate tasks, exchange context, and trigger further ERP actions. However, the SAP community can experience the possibilities of the SAP BTP and SAP Business AI in two AI experience workshops in April in Heidelberg and June in Salzburg. In cooperation with SAP partner Snap, E3 Magazine is organizing these workshops, which have a strong practical focus on BTP, Clean Core, and SAP Business AI.

The IT company Dynatrace has also recognized that most companies are not yet prepared for this AI change. Without strong observability and clear governance, AI/ERP systems will become increasingly difficult to understand and control. This realization applies particularly to agent-based AI. Each AI agent acts independently based on instructions and inputs—not only from humans, but also from numerous first- and third-party agents. A single customer contact can trigger hundreds of background processes in which AI agents make decisions independently, switch roles, and instruct other agents.

However, every agent should be accountable to a human or a higher-level agent, and supervision should always remain with the human—at least, that is the plan. The explosive growth of agent-based communication can no longer be controlled without observability. The challenge no longer lies in optimizing individual models. The decisive factor is controlling the network of autonomous interactions in real time. According to a recent statement by Dynatrace, observability is the basis for secure, scalable, and controllable agent ecosystems.