The AI Paradox

In Latin America, Incumbents Are Primed to Capture the Lion's Share of AI Value—But Success Demands Strategic Execution and Collaboration

Executive Summary

Imagine a technological revolution where the biggest winners aren't the inventors of groundbreaking tools, but those who masterfully wield them in everyday battles. That's the intriguing paradox unfolding in Latin America's AI landscape today.

While the region may not birth the next global AI powerhouse—constrained by towering barriers like capital demands and talent shortages—established companies stand to reap enormous rewards by integrating ready-made AI solutions into their operations.

Drawing insights from McKinsey's, Hi Ventures and Iconiq State of AI Reports, which surveys adoption trends, one can see that AI is shifting from hype to practical infrastructure, with usage climbing and agents (autonomous AI systems) gaining traction. Projections from UNDP (United Nations Development Program) suggest AI could boost LatAm's GDP by 5.4% by 2030, unlocking over $500 billion in value, especially in services.

Yet, incumbents must grapple with real hurdles, from cultural resistance to proving ROI, while exploring partnerships with nimble startups to blend strengths and accelerate innovation. This article explores why incumbents hold the edge—and how they can turn potential into reality.

Part I: Why Latin America Won't Produce AI Platform Champions

The race to build foundational AI platforms—the "operating systems" of tomorrow, as Andrej Karpathy aptly describes LLMs—has become a high-stakes duel between US and Chinese giants. Platforms like OpenAI's GPT series or Google's Gemini command vast resources, outspending entire regional ecosystems; for instance, US Big Tech invests $18 billion annually in AI R&D, dwarfing Brazil's projected $3.8 billion over the next four years. Latin America, rich in ambition, faces structural roadblocks that make competing at this level impossible.

Talent remains a core bottleneck. The global scramble for AI experts, highlighted by Meta's eye-watering $100 million signing bonuses, leaves LatAm firms at a disadvantage. Regional surveys underscore this: around 41% of enterprises cite technical talent shortages as a primary barrier, per Hi Ventures, amplified by informal economies that limit scalable training (as noted in IMF analyses). Network effects further entrench this, with innovation hubs like Silicon Valley drawing top minds away.

Infrastructure adds another layer of challenge. Training advanced models demands massive computing power and datasets, often beyond reach without global-scale investments. Hi Ventures highlights persistent infrastructure gaps in the region, even with governments like Chile and Mexico beginning to fund AI factories.

Distribution moats, such as OpenAI's ties to Microsoft, solidify this duopoly, pushing LatAm toward application-focused strategies rather than foundational builds. Compounding this, AI startup investment in the region remains minuscule—totaling just $137 million in 2024—far too limited to scale local innovators to global relevance.

Part II: The Incumbent Edge—Leveraging Consolidation for Value Capture

Here's where the paradox shines: the very forces locking LatAm out of platform creation open doors for incumbents to thrive in deployment. Instead of reinventing the wheel, established firms can plug into proven technologies, enhancing their operations with AI that diagnoses issues faster or optimizes processes smarter—often without the risks startups face in building from scratch.

Three key pillars underpin this advantage, each amplifying the others in ways that create lasting competitive moats.

1. Data Moat: Decades of accumulated proprietary information—think a bank's decades of credit histories—become goldmines when fed into AI models. This isn't just more data; it's the foundation for highly accurate predictions and personalization that newcomers can't easily replicate, potentially driving significant GDP uplift as per UNDP estimates.

2. Customer Fortress: Trust amplifies AI. Examples include telecoms optimizing networks and retail personalization, boosting satisfaction. In a region where relationships often define business success, incumbents can leverage longstanding customer loyalty to introduce AI-driven enhancements, such as tailored financial advice or predictive maintenance, without the skepticism that greets unproven startups.

3. Capital Engine: With deeper pockets, incumbents can invest in widespread implementations, like deploying AI across hundreds of sites via cloud platforms such as Amazon Bedrock. McKinsey's State of AI reinforces this: Firms with over $500 million in revenue lead in centralizing AI elements and following scaling practices, enabling multi-site rollouts that smaller players struggle to match.

These pillars compound exponentially: better data leads to sharper AI, which delights customers and generates more insights, widening gaps over time. However, incumbents aren't invincible; they often benefit from hybrid approaches, partnering with startups to infuse fresh innovation, as seen in sectors where agile tech complements established scale.

Case Study: Nubank as a Cautionary Exception

Nubank's meteoric rise in fintech—disrupting traditional banks with user-friendly services amid low satisfaction scores and regulatory shifts like open banking—serves as a reminder that complacency can create openings. Yet, this success is outlier territory, tied to a uniquely vulnerable sector - an oligopoly with absurd interest rates and terrible NPS.

In broader fields like insurance or manufacturing, similar breakthroughs are rare, underscoring how incumbents who actively deploy AI can fortify their positions. Hybrids here shine too: Tools like Darwin AI help SMBs but often thrive through alliances with larger players.

Part III: LatAm Context Amplifying Incumbent Strengths

Latin America's business fabric—woven with personal relationships, intricate regulations, and uneven capital access—tilts the playing field toward incumbents. In a culture where trust trumps tech alone, established firms can enhance bonds with AI, rather than disrupt them.

Regulatory hurdles, inspired by global frameworks like the EU's, favor those with compliance expertise. Talent and infrastructure shortages hit startups harder, while incumbents leverage employer brands and existing systems. McKinsey's insights on organizational structures reinforce this: Larger firms appoint dedicated AI leaders earlier, enabling strategic focus that smaller entities struggle to match.

Part IV: Navigating Implementation—From Inertia to Execution

For all their advantages, incumbents face steep implementation hurdles that can stall progress. Legacy systems resist integration, cultural resistance arises from fears of job changes, and proving ROI demands patience—challenges McKinsey identifies as more strategic than technical, with many firms struggling to pinpoint high-impact use cases.

Skills gaps loom large, requiring upskilling programs, while data quality issues from outdated sources complicate efforts. Even with rising mitigation of risks like inaccuracy and IP infringement, larger firms still grapple with talent shortages (50% need more data scientists) and workflow redesign for EBIT impact.

Yet, incumbents hold tools to overcome these: robust budgets for training, phased pilots to build momentum (like rolling out agents in controlled settings), and established change management experience. Success hinges on committed leadership, clear metrics like productivity boosts, and fostering a culture that embraces experimentation. Human-in-the-loop oversight ensures ethical deployment amid these shifts.

One specific note for CEOs: While you are not required to send your own CEO AI memo as some have done lately, your unwavering commitment is key - and beyond that, as a CEO, you cannot outsource learning AI to anyone. If you haven't done so, incorporate it to your daily life intensely and give the example - and soon people will follow.

Conclusion: Embracing the Paradox for a Brighter Future

In Latin America's AI story, the real magic lies not in creating the platforms, but in creatively applying them to solve regional puzzles—from boosting efficiency in bustling markets to personalizing services in diverse communities. Hi Ventures showcases this through incumbents like Itaú (with 500+ GenAI pilots boosting operations) and Domino's (21% royalties increase via automated receivables), for example.

As adoption climbs towards 40% and beyond, incumbents who thoughtfully navigate the paradoxes—balancing their strengths, collaborating with startups and addressing implementation pitfalls—stand to unlock profound value. It's a journey of adaptation and innovation, one that invites reflection on how proven foundations can propel us toward an exciting, AI-enriched horizon.