SAP bets on AI retraining instead of layoffs in restructuring plan

SAP is betting that artificial intelligence can change jobs without forcing a broad cleanup of the payroll. The German software group said in January 2024 that it would spend about €2 billion on a restructuring plan affecting around 8,000 roles, or roughly 7.4% of its workforce at the time, as it shifted more resources toward AI and cloud growth.

The company’s plan was framed around internal reinvention rather than outright cuts. SAP said it would retrain employees for AI-related work where possible and lean on voluntary redundancy programs for some roles, while expecting to end 2024 with a headcount similar to current levels. Most of the restructuring costs were set to hit in the first half of 2024, and SAP said the programme would add about €500 million to operating profit in 2025 through efficiency gains.

Markets immediately read the move as a productivity play. SAP shares rose about 7% to a record high after the announcement, extending investor optimism that generative AI can lift margins for enterprise software companies even when it reshapes work inside the business. At the time, SAP had more than 105,000 employees, making the 8,000-role programme large enough to matter operationally but still limited relative to the size of the group.

The strategy also reflects how quickly SAP had already moved on generative AI. The company had been experimenting with OpenAI’s ChatGPT before the restructuring plan was announced and had laid out plans to embed generative AI into its products early in 2024. That matters because SAP sits at the center of corporate software stacks used by finance, supply chain, and human resources teams, so any redesign of its own workforce offers a glimpse of how AI may spread through other white-collar employers.

For now, SAP is presenting restructuring as a managed transition: reskill where the new work exists, use voluntary exits where it does not, and keep overall staffing broadly stable through year-end. Whether that holds up will shape how other large employers judge the economics of AI, especially if the promised operating profit gains in 2025 arrive without a deeper round of permanent layoffs.