<\/p>\n
SAP leverages agentic AI to transform legacy ABAP code and prepare businesses for a cloud-first future, signalling a disruptive shift in enterprise system modernisation.<\/p>\n<\/div>\n
SAP is betting that the next phase of artificial intelligence will not just speed up programming work, but help re-engineer the vast body of custom code that still sits at the heart of enterprise systems. In an interview published by SAP News, Sonja Li\u00e9nard, the company\u2019s head of ABAP platform, said agentic AI would become central to how customers modernise legacy applications for the SAP cloud. She described the shift as one that could eventually change the market rather than merely improve developer productivity.<\/p>\n
ABAP, the proprietary language SAP created for business software, has been part of the company\u2019s stack for four decades and remains embedded in its core ERP products. TechTarget describes it as SAP\u2019s main programming language, while SAP\u2019s own documentation traces its role back to the company\u2019s early application software. Li\u00e9nard said the language has survived because it was built for enterprise logic, with security checks, authorisation controls and other safeguards designed into the development model from the start.<\/p>\n
That legacy matters because SAP still has a huge installed base to support. Li\u00e9nard said there are roughly five million registered ABAP developers worldwide and about two million active ones, and she argued that ABAP Cloud is now the bridge between older on-premise environments and SAP\u2019s newer cloud products. According to SAP, ABAP Cloud is intended to support a \u201cclean core\u201d approach, helping customers prepare custom code for migration without stripping out the business logic that makes their systems unique.<\/p>\n
SAP is already folding AI into that process. Li\u00e9nard said the company has introduced tools that can explain code and suggest completions, and in February 2026 it extended its custom code management app with AI features aimed at identifying what code does and what changes are needed to keep it future-ready. She also pointed to SAP-ABAP-1, a specialised model released on the generative AI hub in January 2026 to explain ABAP program code. The company\u2019s wider ambition is to package these capabilities into a service that can be used across system versions.<\/p>\n
The longer-term goal is more ambitious still. SAP says it wants agentic AI to coordinate multiple specialised agents that can assess, explain and modify code, while humans retain final control over quality and security. Microsoft defines agentic AI as systems built around agents that can work together on complex tasks, often through an orchestrator that coordinates them dynamically. Li\u00e9nard said this model could eventually help transform old SAP ECC-based systems into cloud-ready applications, while developers shift towards reviewing AI-generated code, shaping business rules and deciding what the software should do. For now, she said, the technology is still early, but the company expects it to become an important part of ABAP\u2019s future.<\/p>\n
Inspired by headline at:<\/strong> [1]<\/a><\/sup><\/p>\n