Introduction
The potential of artificial intelligence (AI) has reached the boards of Mexican companies, but its value remains unrealized in most companies’ ledgers. According to KPMG’s study, “Panorama of Artificial Intelligence in Mexico and Central America 2025,” nearly half (56%) of Mexican companies still fail to identify the commercial value that AI can bring to their business, while 46% of Central American companies face the same challenge.
Fragmented Strategic Maturity
The report reveals a fragmented strategic maturity. In Mexico, 27% of organizations claim to have a well-defined AI strategy aligned with corporate objectives, while 41% only possess general knowledge about AI without a specific business case. Another 31% expresses interest in implementing AI but lacks clarity on execution.
Felix Moreno’s Insights
Felix Moreno, Director of Digital Lighthouse at KPMG in Mexico, stated, “Although 27% say they have the strategy, only 10% confirm that value with firm metrics. We’re still struggling to confirm that value with solid metrics.” He also warned that while organizational knowledge about AI is growing, confirming its impact remains elusive. There are pilots, dashboards, and isolated cases, but the challenge lies in connecting them to key performance indicators (KPI).
Central American Perspective
Luis Rivera, a KPMG partner in Costa Rica, concurred that while strategies exist, they are not always operationalized. He mentioned, “When we focus specifically on strategy, in Central America, we’re left with a much lower level, which highlights the opportunity.”
Executive Expectations
Executives anticipate applying AI primarily for data-driven decision making (79% in Mexico; 71% in Central America), enhancing customer experience (77% and 76%), and optimizing or reducing costs (70% and 78%).
Data Management and Governance
Although automation (73%) and data management (68%) concentrate a significant portion of AI adoption, the data management and governance framework remains immature and reactive.
Data Bottleneck
Felix Moreno explained that the bottleneck is not AI but data. Many companies prioritize post-control measures, such as verifying data quality and organizing information, while neglecting preventive measures like defining data governance processes. Consequently, they operate in a reactive mode, addressing issues after they arise rather than preventing them during design.
Lessons Learned
Strategic-Metrics Gap
The finding that many companies still fail to translate AI into measurable results leaves three practical lessons. First, the gap between strategy and metrics: although 27% of Mexican companies have a strategy aligned, only 10% confirm that value with firm metrics, meaning enthusiasm has yet to cross from the plan to financial statements.
Data Before Models
Second, data comes before models. The study shows that 59% of Mexican organizations focus on checking data quality, and 56% standardize data presentation. However, only 33% define data governance processes and 30% apply advanced security.
Culture and Collaboration
The third lesson is about culture and collaboration. Forty-one percent of Mexican companies use data in isolation by teams, and 31% do so without central processes, limiting value. Luis Rivera emphasized transitioning from individual assignments to a social and collaborative approach within the organization.
Key Questions and Answers
- Q: What percentage of Mexican companies have identified the commercial value of AI? A: Only 10% have confirmed it with firm metrics, despite 27% claiming to have an AI strategy.
- Q: How mature is the strategic approach to AI in Mexico and Central America? A: In Mexico, 27% have a well-defined AI strategy, but 41% lack specific business cases. In Central America, 46% face the same challenge.
- Q: What are executives’ primary expectations for AI implementation? A: Data-driven decision making, enhanced customer experience, and cost optimization.
- Q: What is the main challenge in AI adoption? A: The bottleneck lies in data management and governance, with companies prioritizing post-control measures over preventive ones.
- Q: What lessons can be learned from the study? A: The strategic-metrics gap, prioritizing data before models, and the need for improved culture and collaboration.
