AI readiness in Latin America survey report

Executive summary

Artificial intelligence (AI) has moved from experimentation to a firmly embedded part of business strategy and execution across Latin America (LATAM). In RSM’s survey of 461 business leaders, around 40% of organisations say they are already implementing AI. Among current adopters, 68% plan to scale within the next 12 months. Yet organisational readiness is not keeping pace. Many organisations are attempting to expand with limited experience, incomplete governance, and uneven workforce preparedness.  

“The main risk is not a lack of interest in AI, but that many organisations are trying to scale before they have the capabilities needed to turn that interest into real, measurable, and sustainable value.”

José Gregorio Argomedo

Managing Director, Consulting

RSM

The most important finding is the gap between ambition and institutional readiness. Only 18% of respondents report formal governance frameworks and 17% have employee policies in place, while just over half are still developing them. Only 4% report comprehensive training maturity and just 8% of employees feel fully prepared. This indicates that, for many organisations, the pressure to scale is running ahead of the structures required to do so consistently, responsibly, and at pace.

Our report also identifies a clear relationship between governance and risk discipline. Organisations with formal governance and policies are far more likely to have completed meaningful risk assessments than those without them. This suggests that governance is not an administrative add-on. It is a practical enabler of scale.

At the same time efficiency is the leading adoption driver, cited by 67% of respondents, and investment is concentrating in IT, operations, and marketing. However, functions central to workforce readiness and control, including HR and security, are comparatively less prioritised. As organisations move from pilots to broader deployment, the next phase will depend less on appetite and more on the ability to close three components of the readiness gap: rules, skills, and risk.

These components are closely connected and become more important as AI moves increasingly into day-to-day operations.

1. Rules: Governance and employee-use policies are not yet keeping pace with adoption. Without clear ownership, approved use cases, and usage standards, middle-market firms may find AI harder to control, especially where oversight sits with a smaller leadership bench.
2. Skills: Workforce readiness remains thin. If employees are expected to use AI before they are properly equipped, lean teams may see uneven delivery, inconsistent decisions or control issues emerge quickly.
3. Risk: Risk maturity appears to follow governance maturity. Where rules are clearer, risk assessment is more systematic; where governance is weak, organisations may struggle to scale AI with confidence, trust, and accountability. 

Five numbers that define the readiness gap: 

Methodology note:

This report is based on a 461-response survey conducted in 2025, with approximately 91% of responses coming from Latin America. The sample is leadership-weighted and non-probabilistic for convenience. 

1. Adoption momentum is real, but scale is arriving faster than maturity

AI adoption has progressed beyond curiosity. Around 40% of respondents said their organisations are already implementing AI solutions. More significantly, 68% of adopters plan to scale within the next twelve months, including approximately 25% within six months and 42% within six-to-twelve months.

This gives the market its defining tension. A substantial share of organisations intend to scale quickly, yet around 70% of the six-to-twelve-month expansion cohort has less than one year of experience. In practical terms, many organisations are entering the scale phase before they have accumulated the operational learning, governance routines, and internal capabilities that normally support expansion.

This matters because the challenge is no longer whether leaders see a business case. They do. Efficiency is the primary driver of adoption, cited by 67% of respondents, followed by process automation, data analytics, cost reduction, and innovation. The commercial logic is therefore established. The more consequential question is whether organisations can translate visible value into repeatable, governed outcomes.

Our survey suggests that many are not yet fully equipped to do so. Readiness assessments remain concentrated in early stages, and only a minority of organisations describe themselves as fully prepared. The result is a pattern familiar in emerging technology cycles: strong executive intent, visible use cases, and growing investment, but an institutional base that remains uneven.

“For middle-market businesses, this acceleration creates a practical risk: AI initiatives can move from experimentation to wider use before the organisation has learned how to integrate them. The result is often a collection of promising pilots rather than repeatable business capabilities. Where experience is limited, leaders may need to prioritise fewer, better-governed use cases that are linked to measurable outcomes, rather than allowing adoption to spread informally across the business.”

José Gregorio Argomedo

Managing Director, Consulting

RSM

2. Governance, policies, and risk management remain the clearest readiness gap

The survey’s clearest warning signal sits in governance. Only 18% of organisations report formal AI governance frameworks. Employee policies are similarly underdeveloped: 17% have formal policies in place, while 51% are still developing them. This means a large share of organisations are advancing with incomplete guardrails around acceptable use, decision rights, accountability, and oversight.

These gaps matter because AI introduces risks that are not easily managed through informal practice alone. Questions of data use, model behaviour, content quality, explainability, security, and regulatory exposure all become harder to manage once deployment moves beyond isolated experimentation. The survey findings also suggest that partial risk assessments are more common than comprehensive ones, reinforcing the view that many organisations remain in transition rather than in full control.

One of our survey’s standout findings is the close association between governance strength and risk maturity. Organisations with established governance and formal or developing policies are far more likely to have completed partial or comprehensive risk assessments. In the most mature groupings, risk assessment levels rise to approximately 90% to 96%, compared with materially lower levels among organisations with no plans or no policies, which is to say that governance is strongly associated with better risk discipline.

The takeaway here is that governance should not be treated as a compliance exercise to be added later. It is an enabling structure for scale. Where governance is absent, expansion is more likely to create inconsistency, rework, and avoidable exposure. Where it is present, organisations are better placed to make decisions with confidence.

“This is particularly important for middle-market organisations, where accountability structures may be less layered and specialist risk capacity more limited. In that context, informal or inconsistent use of AI can quickly create exposure around data, intellectual property, model outputs, security, and compliance.”

José Gregorio Argomedo

Managing Director, Consulting

RSM

3. Workforce readiness is lagging behind leadership ambition

Our survey suggested that people readiness is even less mature than governance readiness. Only 4% of organisations report comprehensive AI training maturity. Fewer than one in five say their current programmes are adequate, and just 8% of employees feel fully prepared for AI implementation.

This creates a structural disconnect between organisational ambition and day-to-day execution. In many cases, leaders are discussing expansion while the workforce remains only somewhat prepared or not prepared at all. The survey also suggests that organisational readiness reviews do not consistently incorporate workforce readiness, reinforcing the risk that scaling plans underestimate the human and operational effort required to sustain adoption.

Our findings do not support a simple narrative of job replacement. Respondents are more likely to expect changes in capability than outright workforce reduction. Around 53% expected AI to improve skills, compared with 15% who expect workforce reductions. Reassignment of responsibilities and the creation of new roles also appear in the findings, suggesting that the more immediate effect is job redesign rather than job elimination.

For organisations, this shifts the issue from technology deployment to operating model design. Adoption at scale depends on whether employees understand how tools should be used, where judgement still matters, and how performance and accountability will change. Without that clarity, even well-chosen use cases can stall or create uneven outcomes.

“Workforce readiness can be an especially acute constraint. Employees often hold multiple responsibilities, and formal learning capacity may be more limited than in larger organisations. If AI tools are introduced without role-specific upskilling and change management, adoption can remain a scattered individual capability rather than an institutional one. That increases the risk of uneven quality and operational errors, even where the underlying technology is strong.”

José Gregorio Argomedo

Managing Director, Consulting

RSM

4. Investment is targeting value creation, but not always the functions that support sustainable scale

Planned implementation and investment are concentrated in functions closest to technical delivery and operational value. IT and Marketing lead planned implementation, at 46% and 44% respectively, followed by Operations, Customer Service and Finance. Investment intentions follow a similar pattern, with the strongest reported increases in IT, Operations, and Marketing.

For many, this may seem obvious. These functions offer clear near-term opportunities to improve productivity, accelerate workflows, and support content-intensive activity. However, interestingly, the investment pattern also exposes a weakness in how many organisations are approaching readiness. HR and security rank lower as priorities, despite our findings that workforce preparedness and risk discipline are decisive constraints on scale.

The same pattern appears in the barriers that respondents identified. The leading challenges are lack of specialised talent (38%), systems integration (33%), and security concerns (28%). These are not peripheral issues. They point to the fact that capturing value from AI depends not only on use cases but on integration into systems, processes, controls, and ways of working.

For many organisations, the next phase will therefore require a rebalancing of attention. Technical deployment remains important, but it will not by itself resolve the practical bottlenecks that limit adoption. With AI, sustainable scaling depends on additionally prioritising investment in the functions that help embed, supervise, and train on the use of AI across the business.

“Without a clear architecture, a governance model and value metrics, investments in tools, licences, infrastructure and consultancy can become sunk costs or technical debt that is difficult to sustain. Being prepared is not about having the best tool, but about the foundations that transform curiosity into institutional capacity.”

José Gregorio Argomedo

Managing Director, Consulting  

RSM

The takeaway

AI is already reshaping leadership priorities across the region. The business case is sufficiently visible to drive adoption, investment, and expansion plans. But many organisations are trying to scale before governance, workforce readiness, and risk discipline are fully in place.

The issue is not whether AI matters. It is whether organisations can operationalise it in a way that is safe, scalable, and defensible. In that sense, the readiness gap is not a secondary concern. It is the factor most likely to determine which organisations turn early interest into sustained value. 

Contact us

Complete this form and an RSM representative will be in touch.