ARTFEED — Contemporary Art Intelligence

AI's Business Value Depends on Data Fabric Context, Says SAP Executive

ai-technology · 2026-04-22

Irfan Khan, who serves as the president and chief product officer of SAP Data & Analytics, emphasizes that the primary challenge for successful enterprise AI lies in the quality and context of the foundational data rather than the performance of the models. According to a McKinsey survey, by 2025, half of all companies will implement AI in at least three areas of their operations. However, only 9% believe they are ready to merge AI with their data systems. Implementing a data fabric, which encompasses data federation, a semantic layer, and cross-fabric governance, facilitates AI's interaction with business insights. More than two-thirds of organizations utilizing data fabrics report enhanced data accessibility and control, promoting synchronized AI functions across different business areas.

Key facts

  • Irfan Khan is president and chief product officer of SAP Data & Analytics.
  • By end of 2025, half of companies used AI in at least three business functions per McKinsey survey.
  • Only 9% of organizations feel fully prepared to integrate AI with data systems.
  • One in five organizations consider their data approach highly mature.
  • Data fabric is an abstraction layer spanning infrastructure and logical organization for AI.
  • Data fabric requires federation, semantic layers with knowledge graphs, and cross-fabric governance.
  • Over two-thirds of enterprises with data fabrics see improved data accessibility and control.
  • AI without business context can produce technically correct but operationally flawed decisions.

Entities

Institutions

  • SAP Data & Analytics
  • McKinsey
  • Capgemini

Sources