SAP and Oxford Economics have examined the business benefits of AI. Five insights show where AI delivers real added value – and where there are still gaps.

Companies around the world are investing a lot in artificial intelligence (AI) – but where does it actually bring added value and where are there still deficits? To answer these questions, SAP conducted a study together with Oxford Economics, a leading global economic consulting and research firm.

“We commissioned this research to help companies cut through the hype, understand the real business impact of AI, and accelerate their own measurable results,” said Brenda Bown, chief marketing officer for Business AI at SAP.

The study surveyed 1,600 executives in eight markets: Australia, Brazil, China, Germany, India, Singapore, the United Kingdom and the United States. The survey took place between July and August 2025.

Five key findings:

1. AI plays a key role in global business operations

The study shows that companies are investing in a wide range of AI application areas and are striving for positive returns. Participants reported that AI already supports 25 percent of business tasks in their organizations worldwide; This proportion is expected to rise to 41 percent within two years. Most are scaling AI automation and experimenting with generative AI; Very few of those surveyed said they did not want to introduce either technology.

Despite hurdles to broader implementation, 63 percent of companies report finding solutions and moving forward. Only 3 percent believe that AI will never be central to business processes, decision-making or customer offerings. A majority (59 percent) said AI has helped them solve key business challenges.

When asked about prioritizing AI investments, 44 percent described their approach as piecemeal, 32 percent see it as driven by individual departments and 15 percent call it ad hoc. Only 9 percent invest based on a strategic, holistic approach.

2. Companies are seeing returns on AI investments

For the current fiscal year, participants reported average AI spending of over $26 million per company; Investments are expected to increase by 37 percent in the next two years. Compared to other technologies, 49 percent agreed that AI initiatives deliver positive returns on investment (ROI) more quickly.

Accordingly, AI investments are expected to generate a return of 16 percent (average $4.7 million) this year; In two years, this figure is expected to rise to almost 31 percent (US$12.3 million).

In fact, 79 percent of participants achieved positive returns within one to three years – a finding that strengthens optimism about the rapid impact of AI. Nevertheless, almost two-thirds (65 percent) were unsure or did not believe that AI would achieve its full potential.

To maximize ROI, respondents primarily cited aligning AI initiatives with business strategy and ensuring data availability and quality. This was followed by the use of best-of-breed platforms and solutions.

3. Low confidence in data readiness for AI

While companies recognize the central importance of data in realizing the potential of AI, many are unsure whether they are ready to integrate data effectively. 71 percent of respondents rated data as very important or critical for AI investment decisions. At the same time, however, a majority expressed doubts about being able to share and integrate data responsibly across business areas (55 percent) or with external partners (60 percent).

Almost two thirds (64 percent) rated their data maturity as sufficient overall; At the same time, they named the areas of law, risk and compliance, finance and human resources as the least data-ready – i.e. those that are often particularly complex and sensitive when it comes to data.

Incomplete or inconsistent data (75 percent), poor data quality (69 percent) and data silos (68 percent) were cited as the biggest hurdles to data maturity.

4. Companies invest in talent to maximize the AI ​​impact

Organizations know that technology alone does not guarantee ROI. Human judgment, creativity and acceptance in everyday work are crucial to unlocking the full potential of AI. Equipping employees with the right skills and framework ensures that AI complements rather than replaces contributions.

Almost three quarters (73 percent) of those surveyed further qualify existing employees (upskilling) or retrain them (reskilling), and two thirds (65 percent) identify roles that can be supported by AI. A majority (57 percent) are redesigning tasks to promote collaboration between humans and AI. At the same time, many companies do not consider their teams to be sufficiently prepared to use AI responsibly – there is a lack of comprehensive training and clear guidelines.

This gap leads to risky decisions: 64 percent said employees use unapproved “shadow AI” tools at least occasionally, and 77 percent expressed concerns about the risks. Shadow AI remains a big issue: More than half of companies report inaccurate results, and many also report data leaks or security gaps.

5. The future of AI is agent-based

More than three quarters of companies worldwide (78 percent) are convinced that agent-based AI has medium to very high potential to transform their operations. AI agents are able to make decisions and carry out tasks independently – without constant human input. Expected benefits are primarily process automation, better planning and decision-making, and greater operational efficiency.

Despite this confidence, only 5 percent of companies feel fully prepared to deploy and scale AI agents; 54 percent classify themselves as partially prepared. To prepare for agent-based AI, 63 percent have piloted agent-based use cases and 56 percent are upskilling internal teams or recruiting new talent. Over the next two years, a return of 10 percent – equivalent to $4.3 million – is expected from investments in agent-based AI.

However, only about half of respondents said that agent-based AI will have a significant impact on strategic planning in the short term and that there is a clear, shared understanding within their organization of what agent-based AI is and can do.

The study makes it clear that the opportunity for companies lies not only in recognizing that there is no turning back – but in strategically investing in AI to unlock sustainable and growing returns. Or, as Brenda Bown puts it: “The results show that when companies combine AI with high-quality data, invest in skills and embed intelligence into their core processes, they achieve a significant return on investment.”