The future of AI is teaching machines how businesses work

The conventional wisdom in AI is that the winners will be the companies building the biggest models and the countries investing the most money in chips and data centres.


SAP CEO Christian Klein believes the next winners may look very different. The German software giant’s chief executive argues that large language models are steadily becoming commoditised and that the next battle will be fought in a less glamorous corner of the technology industry: teaching AI how businesses actually operate.


“The first phase of AI was about infrastructure, chips and large language models,” Klein told TOI on Tuesday during his trip to New Delhi where he spoke to the Prime Minister. “The next phase is different. The question is no longer who has the biggest model, but who can teach AI how businesses actually operate.”


That shift, he believes, could create a significant opportunity for India.
For much of the past three years, the AI conversation has been dominated by companies such as OpenAI and Anthropic, along with the chipmakers powering the boom. Countries have been judged largely by whether they have produced a worldclass foundation model.


By that measure, India has often been portrayed as lagging behind. Klein sees the situation differently. As AI moves beyond chatbots and content generation into the day-to-day running of businesses, he believes a different set of attributes will become valuable. Simply training a model on vast amounts of public information will no longer be enough, instead the AI models will have to be taught how supply chains work, how factories operate, how procurement decisions are made, how inventories are managed and how companies comply with regulations.


“The real value will come from understanding supply chains, finance, manufacturing and regulation, and turning that knowledge into business outcomes,” Klein said.


This is the thinking behind what SAP calls the “autonomous enterprise”. Rather than focusing on AI assistants that answer questions, SAP wants to build AI agents that can perform business tasks. These systems are being trained on decades of enterprise data and business processes collected across industries ranging from manufacturing and retail to utilities and transportation.

SAP believes such agents will increasingly handle activities such as financial closing, inventory optimisation, workforce administration, sourcing decisions and contract reviews. The striking part is how quickly Klein believes the shift is already happening.


“No one at SAP is checking contracts anymore,” he said. “That’s all done by AI already.”


Yet Klein does not see foundation model companies as competitors. “I have no sleepless nights about that,” he said when asked about firms such as Anthropic. “You need both. You need SAP’s AI and an LLM to really apply AI successfully.”


As Klein explains it, an AI model may be able to summarise a report or answer a question, but it does not automatically know how to close a company’s books, optimise a manufacturing line or manage a procurement workflow. That knowledge, Klein argues, sits inside enterprise systems and business processes.

Indian IT fears overblown

The implications extend well beyond SAP. For decades, Indian IT services companies built global businesses by helping organisations implement software and redesign business processes. The rise of generative AI has raised concerns that automation could undermine parts of that model. Klein believes the opposite may happen.


“There is a perception that AI threatens the traditional IT services industry,” he said. “I think the bigger story is how it transforms it.”


According to him, companies adopting AI will increasingly need partners who understand both technology and business operations. “Companies will need partners who understand how businesses run, how processes work and how AI can be embedded into day-to-day operations,” he said. “The future is not just about building models. It’s about teaching those models how to create value inside an enterprise.”


That is one reason Klein believes India may be better positioned for the next phase of AI than many assume.


“A lot of people assume India missed the AI race because the biggest foundation models came from elsewhere,” he said. “As AI moves from research to real-world business applications, India’s strengths become much more important.”


Those strengths stem from decades spent helping global companies run their operations. Indian technology firms, consultants and engineers have accumulated deep expertise in sectors ranging from banking and manufacturing to logistics and healthcare.


In Klein’s view, that experience could become one of the country’s biggest advantages as AI becomes more embedded in business operations. The shift is already changing the skills SAP itself is looking for. Rather than hiring large numbers of traditional software developers, the company is increasingly seeking data scientists, data engineers and AI specialists capable of building and training intelligent agents.


“It’s not about coding features any more. That’s done by AI,” Klein said. “It’s about developing very smart agents and providing them to our customers.”


For young engineers, he believes technical skills alone will not be enough. “The best people are the ones who not only have the technical skills to build an agent, but also understand the business problem they are trying to solve,” he said.

SAP’s India ops

The German firm is also increasingly looking to India as a critical part of their strategy to create AI agents that are adept at running business functions. The company recently opened a second Bengaluru campus and Klein said India is now SAP’s largest development centre globally by headcount. SAP Labs India is about 17,000 strong. More importantly, he said, Indian teams are taking ownership of products and AI innovation.


“There is no one in Germany, no one in the US telling the people what to do,” Klein said. “We really shifted complete responsibilities over.”


The company plans to bring its entire global leadership team to Bengaluru in January. Klein also said SAP is working with Indian partners including TCS and the Adani Group on sovereign cloud initiatives that would keep data, operations and AI development within the country. As businesses move from experimenting with AI to deploying it at scale, Klein believes India will play an increasingly important role in shaping how enterprise AI is built and adopted around the world.