SAP CFO Dominik Asam has a reality check for the enterprise AI hype: the technology won't deliver its full potential until it moves beyond chatbots and coding assistants into the core of business operations. Speaking after the company's second-quarter results, Asam said the real payoff depends on governed data and reliable, cost-controlled tools that can safely run finance and supply chain workflows.
Asam's comments come as companies across industries have poured billions into generative AI, but many are still waiting for broad productivity gains. He noted that most AI 'token' usage today goes to what he called 'low-hanging fruit' like chatbots and coding tools, where a wrong answer is usually easy to spot and contains limited risk. The challenge, he said, is scaling AI into high-stakes areas like financial reporting or supply chain management, where errors can have serious consequences.
The Data Governance Hurdle
For enterprise AI to work in critical functions, companies need more than just a powerful language model. Asam emphasized that governed data—meaning data that is accurate, consistent, and properly managed—is essential. Without it, AI tools can produce unreliable outputs that could mislead decision-making in finance or disrupt supply chains.
This focus on data governance aligns with SAP's own strategy. The German software giant has been building AI capabilities into its enterprise resource planning (ERP) systems, aiming to help customers automate tasks like invoice processing, inventory management, and financial forecasting. But Asam's remarks suggest that many businesses are still in the early stages of preparing their data for such advanced use cases.
The CFO also highlighted the need for cost control. Running large AI models can be expensive, especially when deployed at scale across an enterprise. Companies must balance the benefits of AI with the costs of compute power and data management, a challenge that has led some to reconsider their AI spending. This echoes broader market concerns about the slow payoff from AI investments, as seen in recent analyst downgrades of companies like Workday and Salesforce.
What It Means for Investors
For everyday investors, Asam's comments offer a window into the current state of enterprise AI adoption. The technology is not yet delivering the transformative productivity gains that many had hoped for, at least not in the most critical business functions. Instead, the early wins are in relatively safe, low-risk applications.
This suggests that companies selling AI tools for core enterprise workflows—like SAP, Microsoft, and others—may face a longer adoption curve than some expect. Investors should watch for signs that businesses are investing in data infrastructure and governance, which are prerequisites for more advanced AI use. The recent Microsoft launch of a $2.5 billion AI integration unit for enterprise clients underscores the growing focus on making AI work in real-world business settings.
At the same time, the emphasis on cost control could benefit companies that offer efficient AI solutions or help businesses manage AI expenses. The broader enterprise IT spending environment remains steady, as seen in Ingram Micro's positive outlook, but the pace of AI-driven revenue growth may be more gradual than some bulls anticipate.
Asam's message is clear: enterprise AI is a marathon, not a sprint. The companies that succeed will be those that invest in the foundations—governed data, reliable tools, and cost discipline—rather than chasing the latest model. For now, the low-hanging fruit has been picked, and the real work is just beginning.
