Banco Santander is aggressively scaling its artificial intelligence operations, rolling out AI tools to all 185,000 of its global employees. The immediate expansion marks a critical shift from experimental pilots to full-scale enterprise execution, with the bank targeting over €1 billion in combined revenue gains and cost savings between 2026 and 2028.
Previously limited to around 40,000 staff members, the newly democratized access puts tools like Microsoft Copilot, OpenAI, Anthropic, and Google models into the hands of the entire workforce. The financial impact is already materializing; Santander reported €35 million in business value from AI initiatives in the first quarter of 2026 alone, with projections expected to surpass €200 million by year-end.
Automating Fraud Detection and Customer Service
The bank currently operates over 280 automated process agents handling complex tasks such as credit decisions, know-your-customer (KYC) checks, and fraud detection. In Brazil, AI integration has slashed the time required to process card fraud claims by 95 percent, achieving a 90 percent automation rate while keeping errors below 1 percent.
Customer service channels are also undergoing a massive overhaul. In the UK, voice-channel AI is projected to handle 240,000 routine card inquiries annually - roughly 40 percent of the total call volume. This automation is expected to free up 45,000 hours for human service teams and save customers 26,000 hours of wait time. Similar natural-language systems are currently rolling out across Spain for both Santander and Openbank channels.
Agentic AI in Software Development and Payments
Behind the scenes, Santander's technical infrastructure is becoming heavily reliant on machine learning. Approximately 17,000 employees are utilizing agentic AI tools for software development, resulting in 40 percent of the bank's code being generated by AI as of June 2026. Furthermore, Openbank is using AI models to process roughly 100,000 anti-money laundering alerts annually, reducing investigation times from hours to mere minutes.
"AI gains real power when it becomes embedded in daily customer interactions, analytical work, process acceleration, and overall productivity," noted Ricardo Martín Manjón, Chief Data and AI Officer at Santander. The bank is also pioneering AI-driven payments, becoming the first in Europe to test AI-agent-initiated payments with Mastercard, alongside parallel initiatives with Visa in Latin America.
The Enterprise AI Blueprint
Santander’s aggressive rollout is a definitive signal that the experimental phase of enterprise AI is over. By generating 40 percent of its code via AI and automating 90 percent of fraud claims in key markets, the bank is proving that generative models can deliver hard, measurable ROI in highly regulated environments.
The €1 billion target by 2028 is ambitious, but the €35 million already secured in Q1 2026 shows a clear trajectory. What sets this deployment apart is the strict governance framework - ensuring no customer data is shared externally for model training - combined with a multi-vendor approach that leverages Microsoft, OpenAI, and Anthropic simultaneously. This prevents vendor lock-in while matching the right AI model to the specific operational task, creating a blueprint that other global financial institutions will likely be forced to copy.
