As global investment in artificial intelligence accelerates, new data from KPMG reveals a growing disparity between the amounts enterprises spend on AI and the tangible business outcomes they achieve. The findings come from KPMG’s inaugural Global AI Pulse survey, which sheds light on how organizations are deploying AI and the operational strategies shaping value creation.
Investment Soars but Value Realization Lags
According to the survey, organizations worldwide anticipate spending a weighted average of $186 million on AI technologies in the next 12 months. However, only 11 percent have progressed to deploying and scaling AI agents that deliver enterprise-wide business results. This gap reflects the difference between initial AI adoption and fully integrated, value-generating AI operations.
While 64 percent of respondents report that AI is already producing meaningful outcomes, the term “meaningful” encompasses a broad spectrum—from incremental productivity improvements to transformative operational efficiencies that significantly enhance margins. For most enterprises, the latter remains a work in progress.
AI Leaders Versus the Rest: A Stark Performance Divide
KPMG distinguishes between “AI leaders,” organizations scaling or actively operating agentic AI systems, and others still in early phases of AI use. Among AI leaders, 82 percent confirm that AI delivers substantial business value, compared with 62 percent among their peers. This 20-point gap underscores not only differences in tool adoption but fundamental contrasts in AI deployment philosophy.
AI leaders deploy agents that autonomously coordinate work across departments, make decisions without constant human intervention, surface enterprise-wide insights from operational data in near real-time, and proactively identify anomalies before they escalate. In IT and engineering, 75 percent of AI leaders use agents to speed code development versus 64 percent of others. In operations, especially supply chain management, the gap is 64 percent to 55 percent.
Most organizations have integrated AI by adding tools to existing processes, resulting in incremental gains. In contrast, AI leaders redesign workflows first, then embed AI agents into the new structure, yielding superior returns and competitive advantages over a three- to five-year horizon.
Understanding the True Cost of AI Investment
While the headline figure of $186 million per organization sounds substantial, regional disparities offer deeper insights. ASPAC leads with $245 million on average, followed by the Americas at $178 million, and EMEA at $157 million. Within these regions, countries like China, Hong Kong, and the US show particularly high planned investments.
This spending covers model licensing, compute infrastructure, professional services, integration, and the governance frameworks essential for responsible AI operation at scale. Yet, KPMG notes many enterprises underestimate the costs of operational infrastructure—such as engineering efforts to integrate AI outputs with legacy systems, manage data pipelines, and maintain compliance—often incurring higher expenses than initially budgeted.
For instance, integrating vector databases is critical for agentic AI workflows, enabling real-time retrieval of relevant context from vast unstructured document repositories. Selecting the right providers, embedding proprietary data, and managing updates introduces significant engineering and operational complexity.
Governance: Enabler Rather Than Obstacle
A key insight from KPMG’s survey is the correlation between AI maturity and confidence in managing AI-related risks. Only 20 percent of organizations still experimenting with AI feel confident handling these risks, compared to 49 percent of AI leaders. While concerns about data security, privacy, and risk remain universal, mature organizations operationalize governance frameworks that facilitate rather than hinder AI adoption.
Contrary to viewing governance as a compliance hurdle, effective governance embedded within deployment pipelines—through tools like model cards, automated monitoring, explainability features, and human-in-the-loop escalation—empowers organizations to scale AI agents confidently and safely.
Steve Chase, Global Head of AI and Digital Innovation at KPMG International, emphasizes, “There is no agentic future without trust and no trust without governance that keeps pace. Sustained investment in people, training, and change management is essential for scaling AI responsibly and capturing its value.”
Regional Trends and Their Implications
The survey highlights regional differences in AI deployment velocity and organizational attitudes. ASPAC leads in agent scaling at 49 percent, followed by the Americas at 46 percent and EMEA at 42 percent. ASPAC also shows a higher rate (33 percent) of orchestrating multi-agent systems, indicating advanced AI capabilities.
Barriers vary by region: 24 percent of organizations in ASPAC and EMEA cite leadership trust deficits as a major obstacle to AI agent deployment, compared to 17 percent in the Americas. Since agentic AI systems make autonomous decisions, cultures with centralized decision accountability may resist adoption unless governance clearly defines agent authority and escalation protocols.
Expectations for human-AI collaboration also differ. East Asian respondents anticipate AI agents leading projects at 42 percent, Australians favor human-directed AI at 34 percent, and North Americans lean toward peer-to-peer human-AI collaboration at 31 percent. These variations necessitate localized design approaches for agent-assisted workflows in global deployments.
AI Investment Remains a Priority Despite Economic Uncertainty
Significantly, 74 percent of survey participants affirm that AI will stay a top investment focus even amid potential recessionary pressures. This reflects widespread belief in AI’s strategic role in cost optimization and competitive positioning.
For the 89 percent of organizations still experimenting with AI, the message is clear: accelerating AI deployment is imperative but must be balanced with investments in integration and governance to avoid accumulating technical debt and risk exposure that undermine returns.
In summary, KPMG’s Global AI Pulse underscores that the future of enterprise AI hinges on how organizations transition from isolated AI tools to integrated, agentic systems supported by robust governance and operational redesign. This shift is critical to unlocking the promised margin gains and competitive advantages in the evolving AI landscape.
Fonte: ver artigo original
ChatGPT Fails to Detect 92% of Fake Videos Created by OpenAI’s Own Sora Tool
U.S. Grants G42 Approval to Import Advanced AI Chips, Strengthening UAE’s AI Ambitions
Secure Governance Drives Revenue Growth for AI in Financial Services
Google adds deep research and more file types to NotebookLM
