How BBVA Trained 120,000 Employees to Use AI Every Week — and Made Training a Legal Prerequisite
How BBVA Trained 120,000 Employees to Use AI Every Week — and Made Training a Legal Prerequisite
A virtual HR assistant fielding 34,000 employee queries every month across two countries. More than half the workforce using generative AI tools on a weekly basis. Three hours per week saved per employee on tasks that used to be manual drudgery. These are not projections from a consulting slide deck — they are operating metrics from BBVA, the Spanish banking giant with 120,000 employees worldwide.
What makes BBVA's story distinctive is not just the scale of adoption. It is the mechanism: the bank gates all AI tool access on completion of EU AI Act-aligned training, turning regulatory compliance into the on-ramp for enterprise-wide capability building. Harvard Business Review recognized the approach as a benchmark for corporate AI adoption.
Here is how they did it — and what HR leaders at other large enterprises can replicate.
Pillar 1: Training at Scale
BBVA delivered more than 280,000 hours of AI training across the organization in 2025, anchored by a program called "AI Express." The initiative is not optional enrichment. Completing AI Express — which covers responsible AI use aligned with EU AI Act requirements — is a prerequisite for accessing any generative AI tool inside the bank (BBVA newsroom, April 2026).
The governance logic is straightforward: before an employee can use ChatGPT or Gemini on company systems, they must demonstrate they understand what the tools can and cannot do, how data should be handled, and what the regulatory framework requires. This compliance-by-design approach eliminates the gap between policy documents and actual behavior that plagues most enterprise AI rollouts.
The results speak in volume and velocity. When BBVA launched its Gemini Express training module, 105,000 employees enrolled in the first week — the most widely attended training program in the bank's history (BBVA newsroom, April 2026). The bank expanded its ChatGPT licenses from 3,300 to more than 11,000, and among licensed employees, the daily usage rate hit 83% (American Banker, July 2025).
Aila Jimenez, BBVA's Head of Talent & Culture Transformation, has framed the training-first model as inseparable from the adoption strategy: access follows competence, not the other way around.
Pillar 2: A 90,000-Person Community of Practice
Training alone does not sustain behavior change. BBVA built an internal AI community of practice that now includes more than 90,000 employees — roughly three-quarters of the entire workforce. The community has hosted 20 global events designed to share use cases, troubleshoot adoption challenges, and surface new applications (BBVA newsroom, April 2026).
The community also runs BBVA Bot Talent, an internal competition that drew 315 proposals from over 1,200 employees. These are not hackathon novelties. The bank has identified more than 8,000 AI use cases internally, with approximately 700 rated as strategically important (Financial IT, 2026).
Employees have built over 3,000 specialized ChatGPT assistants tailored to specific business functions — from credit analysis workflows to regulatory document summarization. This bottom-up tool creation, guided by top-down governance, is what distinguishes a living AI culture from a one-time rollout.
Pillar 3: The AI Wizards Network
The connective tissue between corporate training and frontline adoption is a network of approximately 750 "AI Wizards" — employees who serve as internal change agents across departments (BBVA newsroom, April 2026).
AI Wizards are not IT staff parachuted into business units. They are domain experts within their teams who have received advanced AI training and act as peer coaches, helping colleagues identify where generative AI can eliminate repetitive work. Juan Ortigosa, BBVA's Global Head of Workplace, has pointed to this distributed model as critical for moving beyond pilot programs to sustained, organization-wide adoption.
The impact is measurable. Employees using generative AI tools save an average of three hours per week on tasks including report drafting, data summarization, translations, and analysis (American Banker, July 2025). Across the workforce, more than 50% now use AI tools weekly, with ChatGPT averaging 12 days per month and Gemini averaging 9 days per month per employee (BBVA newsroom, April 2026).
The HR Assistant: 34,000 Queries a Month
One of the most tangible HR applications is BBVA's virtual HR assistant, which handles over 34,000 employee queries per month across Mexico and Spain. The assistant draws on a knowledge base of more than 2,500 internal documents to answer questions about policies, benefits, and procedures (BBVA newsroom, April 2026).
For HR operations teams, this is the kind of measurable deflection that justifies the investment in training infrastructure. The assistant does not replace HR business partners — it handles the repetitive, high-volume queries that consume HRBP time without adding strategic value.
What HR Leaders Can Replicate
BBVA's playbook offers three concrete takeaways for CHROs and HR transformation leads at other large enterprises:
Gate access on training, not tenure. Tying AI tool access to completion of compliance-aligned training solves two problems at once: it accelerates adoption (employees are motivated to complete training to unlock tools) and it satisfies regulatory requirements before risk materializes.
Build community infrastructure, not just curricula. Training programs have completion rates. Communities of practice have engagement rates. BBVA's 90,000-member community and 750 AI Wizards ensure that learning translates into daily workflow changes, not forgotten certifications.
Measure usage, not just deployment. BBVA tracks days-per-month usage, weekly active rates, and hours saved — not just licenses purchased. This operational telemetry is what allows the bank to demonstrate ROI and identify where adoption is stalling.
With the EU AI Act raising expectations around AI literacy across the European Union, BBVA has positioned its training-first model as a proactive compliance strategy. The bank gates tool access on EU AI Act-aligned training — converting an emerging regulatory expectation into a competitive advantage. The training gate becomes the adoption accelerator, and compliance readiness becomes the foundation for capability. For HR leaders planning their own AI workforce strategies, the question is no longer whether to train at scale — it is whether your training unlocks anything employees actually want to use.
Sources: