“Vague but exciting.” The words were the understatement of the 20th century, scribbled on the margins of Tim Berners-Lee’s famous 1989 paper in which he effectively invented the World Wide Web. In hindsight, we know this was a revolutionary moment, which made the internet accessible to billions of people worldwide and ushered in an era of rapid digitalization.
Today, we’re at a revolutionary moment of similar proportions. Generative artificial intelligence (AI) has gone from niche technology to a topic of global discussion within little more than a year. It has happened at a critical moment, with the world facing multiple geopolitical, economic, and climate crises.
While these challenges urgently require our attention, global human, environmental, and financial resources are already stretched. Generative AI, however, offers hope that we can address all these competing priorities simultaneously. In other words, we’ll be able to achieve more with less.
The enormous potential of generative AI is widely acknowledged, yet there is one key area where its impact for good has yet to be fully realized. If we apply generative AI to how we run business — as a tool to transform companies, supply chains, and entire industries — we’ll accelerate the evolution of our world economy into one that is more sustainable, resilient, equitable, and prosperous.
Generative AI for business can, for example, help find better and faster solutions to the questions millions of organizations around the world face today. For example:
- What steps do I need to take to make my company carbon neutral?
- How can I improve the availability of critical supplies?
- What can I do to make my business more productive and competitive?
- How do we skill and reskill our workforce to meet the challenges of today and tomorrow?
Relying on the recommendations provided by generative AI for such critical matters, however, requires the underlying technology to be extremely trustworthy — much more so than in the consumer application space.
Building Trust in AI
Trustworthy means, first, that generative AI for business has to be relevant. AI can only be as good as the data it is trained on, and the generic data used for today’s most famous large language models (LLMs) will not help companies address their very granular problems. To provide context-specific proposals, relevant AI for business must train and work with real-life enterprise data.
Second, generative AI for business has to be reliable. The stakes in business can be very high: single decisions can affect thousands of customers, colleagues, and the company’s long-term future. That’s why business AI outputs must be provided with the greatest accuracy and quality. And, while AI “hallucinations” may be entertaining in the consumer world, they’re a no-go in business.
Third, generative AI for business has to be responsible. There is an ongoing discussion about how AI models trained on and working with public Internet data may infringe on privacy and copyright regulations. In the business world, this kind of “gray zone” mode of operation is unthinkable.
For businesses to trust generative AI, they need to be sure their data is handled safely and confidentially. They need to be sure that generative AI tools respect and observe data privacy, data ownership, and data access restrictions by their very design, and that they operate only in areas where explicit consent has been given.
These three “R”s — relevance, reliability, responsibility — are the cornerstones of trustworthy AI for the business world. They are also key to building trust in technology as a tool to tackle the biggest challenges of our time.
A Once-in-a-Generation Opportunity
As a global software company, SAP has made relevant, reliable, and responsible AI a top strategic priority, training and working with real-life business data based on the explicit consent of thousands of customers. By design, it follows the access and privacy settings already built into SAP databases and software.
The ethical implementation of AI is ensured by clear guiding principles, internal governance structures, and an advisory panel of external experts. Most importantly, SAP is pushing the quality of generative AI results to be not just “good enough,” but to have the integrity and quality customers expect when they make consequential business decisions.
I believe there is also a once-in-a-generation opportunity on a larger scale: nations and regions that pioneer trustworthy AI for business will see a much faster and broader adoption of generative AI across companies and industries. They will reap the benefits of greater competitiveness, resilience, and sustainability. And they will contribute immensely to a better running world — much like the World Wide Web did three decades ago.
Christian Klein is CEO and a member of the Executive Board of SAP SE.