Harvard Business Review Recognizes BBVA as a Benchmark for Corporate AI Adoption

The ‘shadow AI’ phenomenon, or employee use of AI tools like ChatGPT, Gemini or Claude without informing their technology or compliance departments, risks leaks of confidential data or sensitive company information. However, this concealed use is widespread: according to a survey by MIT Media Lab, 90 percent of surveyed professionals report using AI tools for work tasks, but only 40 percent of companies have purchased official licenses.

In light of this scenario, a team of experts and scholars from BBVA, the London School of Economics, Carlos III University of Madrid, the University of Alicante, and IE University published an article in Harvard Business Review analyzing how a broad, structured adoption strategy can facilitate the integration of this technology into the company and channel its use securely. With an adoption strategy based on facilitating access and encouraging participation, the bank has empowered its employees and decentralized internal innovation, serving as an example to other companies of how to take advantage of existing talent within the organization.

Just two years after launching this strategy, all employees now have access to generative AI tools, and over half of them use these tools on a weekly basis. Furthermore, the bank encourages internal innovation initiatives like BBVA Bot Talent, a competition where employees develop AI solutions for real corporate challenges.

In the early days of adoption, the speed and involvement of senior management were vital in streamlining the strict risk assessment, legal review and General Data Protection Regulation (GDPR) compliance processes in just two months. The Group was therefore ready in record time to sign an initial agreement with OpenAI, allowing the bank to roll out ChatGPT Enterprise within an exclusive, secure cloud for the company, and later to integrate Google Cloud’s Gemini.

Other critical factors included:

• Training for senior leadership. The bank trained 250 of its main directors, including the CEO and the Chair. The goal was for them to view this technology as a valuable assistant, prevent skepticism due to a lack of understanding, and lead by example.
• Adoption through appeal. The limited number of generative AI licenses initially created genuine demand.
• Decentralized innovation. Instead of leaving solution creation in the hands of a centralized team, employees who best understood business processes were empowered. To do so, an adoption network was created, combining ongoing insights, and introducing the roles of champions (leaders) and wizards (local experts), who were tasked with identifying the most valuable use cases.
• Human-controlled governance. Empowerment is combined with oversight. AI does not directly write in central databases without human validation, and the quality of the assistants is regularly evaluated to ensure transparency and security.

As the article published by Harvard Business Review, a publication considered a global reference in management, strategy and leadership, concludes, BBVA’s experience can serve as an important lesson for companies: success does not come from imposing top-down technology guidelines, but by trusting the innovators already within organizations and allowing them to be the ones to help unlock their full potential.

Ultimately, the true success of AI does not lie in rolling out cutting-edge technology, but in how people use it to fundamentally transform the way they work and improve customer service.