For the second year running, a global study reveals that while enterprises are increasingly successful in moving AI projects into production, the return on investment has not improved significantly. The rise of agentic AI brings new challenges, emphasizing the need to close the ‘last mile gap’ and tighten governance to unlock AI’s full value.
- 93% of enterprises improved AI production over the past year
- 57% report ROI growth no faster than their AI investments
- Governance advances accelerate AI delivery and trustworthiness
What happened
The Fifth Annual Domino Enterprise AI Report surveyed 639 AI leaders in North America, the UK, and Europe, revealing that most enterprises have enhanced their ability to move AI from experimental phases to production environments. A significant 93% of respondents noted improvement in deploying AI at scale within their organizations, with over half reporting substantial advancement in production capabilities.
Despite these operational gains, the financial returns on AI investments have not kept pace. More than half of the enterprises reported that their AI ROI is growing at the same rate or slower than their expenditures. This stagnation is observed globally, with a majority of respondents in the US, UK, and Europe expressing that AI-related costs are outstripping the financial benefits derived from these initiatives.
Why it matters
The continuing shortfall in ROI despite widespread AI deployment raises concerns about the effectiveness of current AI strategies and expenditure controls. The report highlights the emergence of a critical ‘last mile gap’—the difficulty in translating AI production models into business-ready applications that deliver tangible value. This gap threatens to undermine enterprise confidence and the strategic benefits AI can offer.
Additionally, agentic AI, which operates with more autonomous decision-making capabilities, introduces new governance challenges. Enterprises with rigorous governance frameworks demonstrate faster, higher-quality AI deployments. Conversely, organizations lacking robust oversight risk regulatory sanctions, revenue losses due to lagging innovation, and increased scrutiny from boards, especially as autonomous AI systems become more prevalent.
What to watch next
Enterprises will need to focus on closing the ‘last mile gap’ by integrating AI production models into secure, scalable applications accessible to business users. Accelerating this process could be aided by advancements in coding assistants that reduce development times but organizations must remain vigilant about the risks that incomplete governance can bring.
Governance remains a critical front for competitive advantage in AI. Companies investing early in comprehensive policy frameworks, audit trails, and compliance mechanisms for agentic AI are expected to outpace peers in delivering impactful AI solutions. Monitoring how enterprises balance innovation with oversight in the coming year will provide key insights into the maturing AI economy and its tokenomic realities.