Companies worldwide are experiencing a significant jump in AI return on investment, with expected returns reaching 21% this year compared to 16% last year, according to new research from SAP and Oxford Economics. The average global business now spends $28 million annually on AI and expects to generate $6.3 million in returns, a figure projected to nearly triple to $15.9 million within two years.

This marks a turning point for enterprise AI. After years of pilot projects and experimental deployments, businesses are moving from testing to execution, and the financial results are becoming visible. The shift is particularly pronounced in agentic AI, a category of AI systems designed to operate with minimal human intervention. Average ROI from agentic AI is expected to reach $17.6 million in two years, more than quadrupling from last year's estimates of $4.3 million.

The research surveyed 2,600 business leaders across 13 countries, including Australia, Brazil, Canada, China, France, Germany, Italy, India, Japan, Singapore, Thailand, the United Kingdom, and the United States.

Why Are Companies Finally Seeing Real Returns?

The jump in ROI reflects a fundamental shift in how organizations approach AI. Nearly a third of all tasks in the average business, about 30%, are now supported by AI, with that figure expected to climb to 48% within two years. More importantly, strategic investment in AI has nearly doubled year-on-year to 17%, suggesting that companies are moving beyond scattered, ad-hoc AI experiments toward coordinated, enterprise-wide strategies.

Business leaders are also increasingly optimistic about AI's potential. Over 69% of businesses report satisfaction with their current AI ROI, even though more than two-thirds acknowledge that AI is not yet achieving its full potential. The optimism is partly driven by expectations around agentic AI; over 83% of businesses believe agentic AI has moderate to very high potential to transform their organization.

Beyond financial metrics, companies are rethinking workforce strategy. A separate survey from commercial real estate firm JLL found that 60% of senior business leaders expect their organizations to increase headcount over the coming years, while the same proportion believe AI will reinvent existing roles rather than replace them. This challenges the widespread narrative about AI-driven job losses. Organizations further advanced in AI adoption are more likely to recruit full-time employees, invest in entry-level talent, and redesign jobs so people work alongside AI systems.

"AI has moved from experiment to execution, and that's beginning to show real returns. But there's still a long way to go. Because AI that lacks context, whether that's processes, data, or governance, at best creates activity without outcomes and at worst creates risk," said Sean Kask, Chief AI Strategy Officer at SAP.

Sean Kask, Chief AI Strategy Officer at SAP

What's Blocking Companies From Capturing Even More Value?

Despite the positive ROI trends, significant obstacles remain. The SAP research identified five critical barriers that are slowing broader AI adoption and limiting potential returns:

• Data Quality Issues: 73% of companies report challenges with incomplete data, and 79% experience rework, delays, or backlogs due to low-quality AI outputs.
• Workforce Readiness Gaps: Nearly 78% of businesses are unsure or agree that their company upskilling is not keeping up with AI tool evolution, while 69% report shadow AI use happening at least occasionally.
• Governance Deficiencies: Only 12% of businesses say their skills or processes and frameworks are fully ready to govern AI effectively, creating significant risk exposure.
• Leadership Vacuum: Under half of companies have a dedicated AI leader responsible for adoption, clear frameworks for AI development, or training on AI capabilities and risks.
• Agentic AI Preparedness: Only 3% of businesses say they are fully prepared for agentic AI, while the majority are either partially prepared or not prepared at all.

The governance challenge is particularly acute as companies scale agentic AI. Today, 38% of companies do not have a human-in-the-loop process for agentic workflows, 37% lack permission and access controls for agents, and only 44% maintain a registry of the agents deployed across their business. More than two-thirds of businesses either agree or are unconvinced that they are deploying agents faster than they can govern them.

Skills shortages also loom large. The JLL survey found that 36% of respondents cited gaps in AI, analytics, and emerging technology skills as their biggest challenge over the next three to five years. Other barriers included limited change management expertise, organizational silos, and difficulties measuring the impact of AI investments.

How to Build a Sustainable AI Strategy That Delivers Real Value

To move beyond isolated wins and unlock enterprise-wide AI value, organizations need to address several interconnected areas:

• Integrate AI with Context: Connect AI systems to the data and processes that run your organization, ensuring AI has the context and governance to drive trusted results rather than creating activity without outcomes.
• Establish Clear Governance Frameworks: Develop human-in-the-loop processes for agentic workflows, implement permission and access controls for AI agents, and maintain a registry of all agents deployed across the business.
• Invest in Workforce Development: Prioritize upskilling programs that help employees work alongside AI systems, and design roles that leverage human judgment and creativity in partnership with AI capabilities.
• Appoint Dedicated AI Leadership: Assign clear accountability for AI adoption strategy, develop frameworks for AI development, and ensure training on AI capabilities and risks reaches all relevant teams.
• Improve Data Quality: Address incomplete data challenges that lead to rework and delays, as data quality remains the biggest obstacle to AI ROI.

The UK financial services sector offers a useful case study. The government's Financial Services AI Adoption Plan acknowledges that while the sector has achieved 75% adoption rates among regulated firms, significant further opportunities exist. The plan emphasizes that scaling AI requires moving beyond isolated pilots to rapid, sector-wide deployment while maintaining trust and financial system resilience.

"Realizing real value from AI is not going to be easy because it demands a new approach. Businesses large and small will need to connect AI to the data and processes that run their organizations, and make sure it has the context and governance to drive trusted results. That's what we call the Autonomous Enterprise. This isn't a technical change; it's a human one. Because you can only achieve real value if agents, processes, and people work as one," noted Sean Kask.

Sean Kask, Chief AI Strategy Officer at SAP

What Does the Future Look Like for Enterprise AI?

The trajectory is clear: AI is moving from novelty to necessity. The fact that 69% of businesses are satisfied with their current AI ROI, combined with expectations that agentic AI ROI will quadruple within two years, suggests that companies are past the point of questioning whether AI delivers value. The question now is how quickly they can scale responsibly.

However, the gap between adoption and readiness remains substantial. While 78% of business leaders believe AI will significantly influence their real estate and workplace strategy, only 31% are actively preparing workplaces to support collaboration between people and AI, and just 15% have reached what JLL describes as the optimization stage of AI adoption. The majority remain in the planning or monitoring phase.

The organizations that will capture the most value are those that treat AI adoption as a holistic transformation, not a technology implementation. This means aligning data infrastructure, governance processes, workforce development, and leadership accountability around a shared vision. The financial returns are real and growing, but only for companies willing to do the harder work of building the organizational foundations that make AI trustworthy and effective.