By Luis Gonçalves*
In Mexico and Latin America, the adoption of AI is advancing faster than the trust and fluidity of collaborators in organizations. This gap gives rise to a phenomenon known as shadow AIa shortcut that expands the attack surface and strains corporate data management. The strategy to combat it is not only about increasing controls, but also about winning trust with responsible AI, protected data and superior experiences.
Generative AI is today a vector of transformation, but also of risk if it is adopted without trust or clear direction. In Mexico, 60% of employees admit they do not fully understand how to implement GenAI securely, while 94% of organizations expect GenAI to strengthen their security operations (1): high expectations coexisting with uneven capabilities.
When the “official” experience does not accompany, the shadow AI: collaborators who use their own AI tools (with personal accounts or their own devices) to solve real needs, creating security, traceability and compliance gaps. The problem is no longer just technical: it is one of trust, clarity and user experience.
Shadow AI: when profit surpasses corporate management
He shadow AI It is not born from whim, but from an equation that favors immediate usefulness: if the corporate solution takes time or is confusing, people look for a shorter path.
Deloitte shows that even as companies promote the use of AI and offer training, many employees (69%) continue to use personal tools at work. Because? Because those apps solve tasks with less friction.
The problem is the hidden cost: credentials outside the perimeter, sensitive data without controls and results that are impossible to audit—exactly where Mexico shows the greatest sensitivity. In the country, almost all organizations face obstacles to integrating security into the strategy, while a majority express concern about their ability to manage devices and protect data.
Added to this is the pressure to advance in AI with ready and secure databases, a point where the most cited challenges are privacy, the processing of sensitive information and data preparation. In other words: the shadow AI It thrives where corporate expertise is less useful than the personal alternative, and right on the fronts (data, privacy, threat response) where we have the least margin for error.
Towards AI that inspires confidence
As John Scimone, president and chief security officer of Dell Technologies, mentions, “Trust is the foundation of AI innovation, and without it, companies hesitate to fully unlock the potential of AI.” In the case of Mexico, 97% of organizations face barriers to adopting GenAI, with data security and lack of internal experience being the main obstacles (2). In this context, a strategy that combines governance, training and clear experience not only mitigates risks: it turns corporate AI into a more useful, safer and more reliable solution than any informal alternative.
In this sense, according to the findings of its latest study on generative AI in Latin America, Accenture suggests that trust is not a state, but an active practice. Among the five strategic imperatives it proposes for responsible adoption, the fourth focuses on closing the trustworthy AI gap. This includes establishing ethical principles and clear governance; assess risks such as bias, privacy and security with continuous metrics; guarantee compliance and permanent monitoring; and transparently communicate data usage and controls available to the user.
Trust is built
For leaders in Mexico, the message is clear: trust in AI is not imposed, it is built. Every day without a punishable and superior alternative is a day of greater exposure. But it is also an opportunity. Organizations that prioritize user experience, data protection, and transparency are charting the path to trustworthy AI. And in that journey, people are the engine of change. With adequate training, clear tools and spaces for participation, teams can become strategic allies in security. Resilience is not achieved with technology alone: it is achieved when people trust, understand and actively participate in its use.
About the author:
*Luis Gonçalves He is president of Dell Technologies for Latin America
The opinions expressed are solely the responsibility of their authors and are completely independent of the position and editorial line of Forbes Mexico.
Follow information about business and current affairs in Forbes Mexico
Source (1) and (2): Dell Technologies Survey 2025
