Deep Read: Wie SAP trotz KI-Rückstand wieder gewinnen will

San Francisco, Palo Alto, and Düsseldorf witnessed SAP's strategic move to regain ground in the artificial intelligence (AI) field in late September. Former SAP technology leader Philipp Herzig recounted a near job loss incident during a supervisory board meeting in early 2022 when a proposal was made to develop a large language model for AI, similar to those from OpenAI, Anthropic, and Google.

Herzig dismissed the idea as foolish, reflecting a long-standing skepticism towards AI at SAP. Today, SAP opts to purchase language models from competitors rather than invest in in-house development, which they consider "mass-market" solutions. Instead, SAP allocates funds to customizing these models for specific business processes or implementing table models that efficiently process business data.

Herzig showcased SAP's latest model from the Freiburg lab, Prior Labs, acquired in the summer. This appearance in California marks a new phase for SAP, a once-valuable German company. The presentation was part of SAP's comeback strategy following years of struggling to respond to the AI boom. While the implementation of AI promises turned out to be more challenging than anticipated, SAP's recent offensive led by CEO Christian Klein shows promising signs.

Customized table models by Prior Labs, a team reminiscent of the company's early years, and the revamped AI assistant Joule are expected to elevate the customer experience to the premium level that SAP's high prices suggest. SAP is investing in novel workshops to prepare personnel for the AI future. According to the Handelsblatt, the leadership under Klein is determined not to waste any more time on AI.