Artificial intelligence (AI) is no longer a technology of the future or a tool that organizations can afford to observe from a distance. On the contrary, its integration into decision-making, risk management, and crisis anticipation presents new challenges for boards of directors and senior management, but also significant opportunities to improve the speed, quality, and competitiveness of companies.

This was one of the main conclusions of the panel “AI and technology as allies in governance and crisis management”, held as part of the annual meeting of the Association of Corporate Secretaries of Latin America (ASCLA), which took place at the KPMG Tower in San Isidro.

The panel was moderated by Willy Guerrero, lead partner of the Consulting area at KPMG in Peru, and included the participation of Claudia Valdivia, independent director and general director of Postgraduate Studies at UTEC; Flavio Corradini, vice president of the MCLAC Enterprise AI Platform at SAP; and Iván Herrero Bartolomé, data director of the Intercorp Group.

One of the main messages from the conversation was that the biggest risk for organizations is being paralyzed by the fear of using AI. “This fear can end up sabotaging any good strategy. If a company decides not to move forward out of fear, there will always be another that will take that risk in a different way, use AI, learn faster, and ultimately gain a competitive advantage,” highlighted Flavio Corradini.

The winning equation: data plus processes and governance

Willy Guerrero proposed a clear formula for organizations to leverage the potential of AI without losing control over its risks. “A fundamental equation for harnessing the potential of AI is combining reliable data, robust processes, and clear governance. Without accurate and reliable data, and without adequate processes, the results will not be as expected. Furthermore, adoption must be supported by clear parameters, a defined risk tolerance, usage guidelines, and properly identified responsible parties,” stated Guerrero, lead partner of KPMG’s Consulting practice in Peru.

Following this line of reasoning, Claudia Valdivia focused on one of the main gaps that organizations currently face: the technology is already in place, but governance has not yet advanced at the same pace.

According to Diligent’s Board Index 2026, cited during the conversation, only 22% of boards have adopted a formal AI governance policy, while the remainder are still discussing the issue or have not clearly incorporated it into their agenda. “AI is already here and being used. The problem is that we often don’t know who the human responsible party is for each system, what protocol is activated if something goes wrong, or who is accountable. We need to assign responsibilities nominally. AI is not autonomous and cannot handle a crisis: there always needs to be a human responsible party,” stated Valdivia.

Fewer errors and a “digital twin” for directors

In turn, Iván Herrero highlighted the potential of AI to anticipate scenarios, test crisis protocols, and complement the analytical capabilities of teams and boards of directors. He explained that, in certain repetitive processes of low to medium complexity, tests showed error rates between 5% and 8% with AI, compared to approximately 10% to 12% in processes performed by people, depending on the sample.

“We can’t simply stop using AI because it sometimes makes mistakes. The important thing is to measure the actual error rate and compare it to our existing processes. In some cases, AI can significantly reduce that margin and, moreover, maintain consistent performance,” Herrero explained.

Another concept that generated great interest during the panel was that of a “twin director,” or a kind of digital counterpart or AI agent that accompanies board members in analyzing information and scenarios.

Herrero even mentioned the future possibility of a “director agent”: an AI tool that doesn’t replace people or make decisions, but helps detect blind spots, identify variables that haven’t been considered, and question the analysis before a high-impact decision is made.

The panel conveyed a common message: the discussion should no longer focus on whether organizations will use artificial intelligence, but rather on how they will incorporate it responsibly, competitively, and in line with their business objectives. To that end, experts agreed on the need to combine experimentation and agility with clear governance, assign human responsibility, and classify use cases according to their level of risk and impact. AI can expand the analytical capabilities of the Board, but critical judgment and responsibility for decisions should always remain with people.