SKN CBBA - ...
SKN CBBA
Cross Border Banking Advisors
SKN | BNP Paribas Scales Generative AI Across Digital Banking and Mortgage Services

Finance

SKN | BNP Paribas Scales Generative AI Across Digital Banking and Mortgage Services

By Or Sushan

โ€ข

November 18, 2026

Key Points

  • BNP Paribas is expanding its generative AI virtual assistant, Genius bar, across its retail banking operations, targeting broader deployment to customers by November 2026.
  • The assistant handled 84,600 conversations per month by May 2026, while its escalation rate to human customer service declined from 66% to 38% over one year.
  • AI-powered mortgage processing has reduced document review time and enabled the bank to issue mortgage offers within two hours of financing approval when combined with its digital mortgage journey.

BNP Paribas Expands AI-Powered Customer Support

BNP Paribas is expanding its use of generative artificial intelligence across everyday banking and mortgage services, seeking to improve customer experience while increasing operational efficiency.

The initiative responds to changing customer expectations as more banking interactions move to digital channels. In 2024, the bankโ€™s approximately 7.7 million clients generated nearly one billion digital interactions, 40 million remote connections and 1.5 million in-branch appointments.

To manage this demand, BNP Paribas has adopted an โ€œoptichannelโ€ model that positions digital services as the primary entry point for customer acquisition, sales and self-service, while retaining human support through its Client Service teams and branch network.

Two initiatives have received recognition in The Digital Bankerโ€™s 2026 Global AI Innovation Awards: Genius bar, the bankโ€™s generative AI virtual assistant, and its AI-powered mortgage journey combining intelligent document processing with personalized mortgage rates. The projects received awards for Best AI Virtual Assistant โ€“ France and Outstanding Use of AI in Customer Personalisation โ€“ France, respectively.

Genius Bar Brings Generative AI Into Everyday Banking

Genius bar was developed to provide immediate, personalized responses to routine banking questions while allowing customers to access human assistance when necessary.

The assistant is integrated into HelloรฏZ for Hello bank! and is being introduced through the Telmi assistant for BNP Paribas. It supports questions about daily banking, financial products, offers and pricing, with the ability to transfer more complex interactions to Client Service.

The rollout began with a pilot involving 10,000 Hello bank! customers in June 2025. By January 2026, the service had expanded to the bankโ€™s entire Hello bank! customer base of more than one million. BNP Paribas began its wider rollout in April 2026 and is targeting deployment to all customers of its brand by November.

The technology combines large language models, retrieval-augmented generation and conversational AI. The models are hosted on BNP Paribas infrastructure, with Mistral AI supplying the language models and Giskard supporting AI security and testing.

By May 2026, Genius bar was handling 84,600 conversations per month, comprising approximately 147,500 messages. Average response times were below seven seconds, and the system could recognize 25 transaction intents that customers could complete independently.

One important operational indicator is the reduction in interactions requiring human intervention. The escalation rate to Client Service fell from 66% in May 2025 to 38% in May 2026.

Bertrand Cizeau, Head of Hello bank! and Deputy Head of Individual and Entrepreneur Banking at BNP Paribas, said the transition from predefined machine-learning responses to generative AI enables more tailored interactions and represents a significant development in the bankโ€™s customer service model.

AI Accelerates Mortgage Document Processing

BNP Paribas is also applying artificial intelligence to mortgage services, where lengthy documentation and approval procedures can create delays for borrowers.

The bank granted approximately 37,000 residential mortgage loans to individual customers in 2025. In France, the approval process typically takes between 45 and 60 days after a property-sale promise is signed.

Its approach combines Intelligent Document Processing (IDP) with AI-driven interest-rate personalization. Together, these technologies are designed to streamline document validation and improve the alignment of mortgage pricing with individual customer profiles.

The IDP system reads and validates mortgage documents within seconds, generating control reports that identify discrepancies between information entered into the bankโ€™s internal systems and the documentation supplied by customers.

This automation saves relationship managers approximately 30 minutes per mortgage file. Previously, the same process required between 1.5 and two hours.

Since January 2026, approximately 124,000 documents had been reviewed, with 21.1% validated without human intervention. The figures demonstrate the potential for automation to reduce administrative workloads while allowing employees to focus on cases requiring additional assessment.

Personalized Mortgage Rates and Faster Offers

The second component uses an AI-driven model to personalize mortgage rates according to customer characteristics and the potential long-term value of the relationship.

The model considers customer profiles, transaction types and segmentation, while incorporating projected eight-year net banking income. It uses approximately 40 parameters to estimate future client value and determine an individualized mortgage rate.

Since June 2026, the combination of these two solutions with BNP Paribasโ€™s digital mortgage journey has enabled the bank to issue a mortgage offer within two hours of financing approval. The period between acceptance of the preliminary offer and delivery of the final loan offer has also been reduced by 10 days.

Alban Marty, Head of Individual and Entrepreneur Banking at BNP Paribas, said the objective is to make the mortgage process faster and more transparent while preserving the human guidance required for an important financial decision.

Catherine Mathon, COO of BNP Paribas Commercial and Personal Banking in France, emphasized that the bank is prioritizing AI initiatives according to measurable benefits for customers and employees. She cited the reduction in processing time achieved through intelligent document processing as an example of the practical impact.

Closing Insights

BNP Paribasโ€™s initiatives demonstrate how generative AI and document automation are moving into core retail banking operations. The Genius bar rollout focuses on handling routine customer interactions more efficiently, while mortgage automation targets a process where administrative delays can directly affect customers and employees.

The next stage will depend on expanding both systems while maintaining accuracy, security and appropriate human oversight. The bank also plans to broaden the range of documents supported by its IDP solution. Adoption rates, customer satisfaction, processing efficiency and the reliability of personalized pricing will be important indicators of whether these technologies can deliver sustainable operational benefits at scale.

For financial institutions, the broader strategic consideration is how to integrate AI into existing service models without weakening customer relationships, data protection or decision-making standards.

ย 

For a confidential discussion regarding retail banking strategy, insurance distribution models, customer loyalty ecosystems, digital financial services, or cross-border financial innovation opportunities, contact our senior advisory team.

Leave a Reply

Your email address will not be published. Required fields are marked *

More like this

Seraphinite AcceleratorOptimized by Seraphinite Accelerator
Turns on site high speed to be attractive for people and search engines.