BNP Paribas AI Cloud Deal: A Powerful New Shift
Crazyrites Editorial Team
September 25, 2026
Introduction
BNP Paribas has announced a new five-year partnership with Google Cloud as the French banking group expands its use of cloud technology and artificial intelligence. The BNP Paribas AI partnership will give the bank access to Google Cloud infrastructure, Gemini models and Gemini Enterprise.
However, the agreement does not mean that all BNP Paribas data will move to the public cloud. The bank plans to keep certain highly sensitive customer information and critical operations outside public cloud infrastructure. This approach shows how large financial institutions are trying to use AI while keeping strong controls around important data.
What the New Partnership Includes
The BNP Paribas AI partnership is designed to expand the bank’s cloud and AI capabilities over the next five years.
Under the agreement, BNP Paribas will have access to Google Cloud’s AI-focused infrastructure, Gemini models and Gemini Enterprise. The bank plans to use these technologies for different business needs rather than moving every workload to one platform.
The partnership also supports BNP Paribas’ existing multi-cloud and multi-model strategy. In simple terms, the bank wants to use different cloud platforms and AI models depending on the type of work involved.
This can give the bank more flexibility while allowing it to keep tighter control over sensitive workloads.
AI Use Will Start With Banking Workflows
One of the first areas covered by the BNP Paribas AI partnership is Corporate & Institutional Banking, also known as CIB.
BNP Paribas plans to use AI agents for selected workflows. One example is helping teams prepare corporate credit memos. Other possible areas include sales, trading, research and structuring.
Gemini models are also planned for integration into LLM@CIB, the bank’s internal generative AI assistant. BNP Paribas says the platform is already available to more than 65,000 employees.
The idea is not simply to give employees a general chatbot. Instead, the bank wants to build AI tools for specific business tasks, with access and controls designed around each use case.
Sensitive Customer Data Will Stay Outside Public Cloud
The most important part of the BNP Paribas AI partnership is the bank’s decision to maintain limits around public cloud use.
BNP Paribas has said that its existing security and data governance rules determine which data and workloads can be processed in a public cloud environment. Certain categories of data will not be stored there.
Reuters reported that medical information held by the bank’s insurance operations is an example of highly confidential data that BNP Paribas avoids putting into public cloud systems. The bank will also continue using its own on-premises infrastructure for certain operations.
This means the agreement with Google Cloud is not a complete replacement of BNP Paribas’ existing technology systems.
Instead, the bank is taking a mixed approach.
Why This Matters for Banking
The BNP Paribas AI partnership highlights a wider challenge for the financial industry.
Banks handle large amounts of information about customers, businesses and financial transactions. Some of this information can be extremely sensitive. At the same time, AI can help banks process information faster and support employees with routine or complex tasks.
Finding the right balance is therefore important.
A bank may want the computing power and AI tools offered by a major cloud provider, while still keeping the most sensitive information under tighter control.
BNP Paribas’ approach shows how this balance can work in practice.
Security Controls for AI Agents
The BNP Paribas AI partnership also includes controls for AI agents.
According to BNP Paribas and Google Cloud, AI agents will be authenticated and given access only to the resources required for their assigned tasks. Their connections with the bank’s information systems and their activities will also be monitored and controlled.
These controls are important because AI agents can do more than simply answer questions. An agent can potentially work across business systems and perform several steps to complete a task.
For a financial institution, limiting what an AI agent can access can reduce the chance of an unintended action or unnecessary exposure of information.
Data Ownership and AI Training
Another important point in the BNP Paribas AI partnership concerns the use of bank data.
Google said BNP Paribas will retain ownership of its data. Reuters also reported that the data will not be used to train, tune or refine Google’s models under the agreement.
For a large financial institution, such terms can be an important part of an enterprise AI strategy.
Businesses generally need clear rules about where their information goes, who can access it and how it can be used. These questions become even more important when AI systems are involved.
BNP Paribas Will Continue Using Other AI Models
The BNP Paribas AI partnership does not mean the bank will rely only on Google.
Reuters reported that BNP Paribas plans to continue using open-source AI models and maintain its relationship with French AI company Mistral. Different providers can therefore be used for different applications.
This supports the bank’s multi-model strategy.
Rather than selecting one AI system for every task, BNP Paribas can choose technology based on the needs of each business process.
That approach may also help the bank avoid becoming too dependent on one technology provider.
Where Google Technology Could Be Used
The BNP Paribas AI partnership has several possible business applications.
| Area | Planned or Potential Use |
|---|---|
| Corporate banking | AI support for credit memo preparation |
| Trading | AI assistance for selected workflows |
| Research | Support for research-related tasks |
| Sales | AI tools for selected business activities |
| Data management | Support for less sensitive data processes |
| Invoice processing | Automation and AI assistance |
| Employee support | Gemini integration into LLM@CIB |
| AI agents | Purpose-built agents for specific tasks |
The bank has stressed that the technology will be selected according to the needs of each use case and within its security and governance framework.
What This Means for the Future of Banking AI
The BNP Paribas AI partnership is part of a broader move toward enterprise AI in financial services.
Banks are increasingly looking at AI for tasks such as document preparation, research, customer support, data management and risk-related work. However, financial institutions also face strict requirements around security, privacy and control.
This makes the banking sector different from many other industries.
For BNP Paribas, the current strategy suggests that AI adoption will happen step by step. Less sensitive workloads can move toward cloud-based AI, while highly confidential information can remain within more controlled environments.
A Hybrid Approach Could Become More Common
The BNP Paribas AI partnership also highlights the growing importance of hybrid technology strategies.
A hybrid approach allows a company to combine public cloud, private systems and different AI models. This can help businesses match technology choices with security and operational needs.
For banks, this can be especially useful because not every application has the same level of sensitivity.
A simple internal productivity tool may have different requirements from a system handling highly confidential customer information.
Therefore, one technology setup may not be suitable for every banking task.
Conclusion
The BNP Paribas AI partnership with Google Cloud marks a significant step in the bank’s technology strategy, but it is not a complete move away from its existing infrastructure.
BNP Paribas will gain access to Google Cloud infrastructure, Gemini models and Gemini Enterprise while continuing to use other cloud and AI technologies. At the same time, the bank plans to keep certain highly sensitive information and critical operations outside the public cloud.
The partnership shows that financial institutions can approach AI adoption with both innovation and careful data controls. As AI becomes more common in banking, the focus will likely remain on finding practical ways to improve productivity while protecting sensitive information.
For customers and businesses, the key issue will be how banks manage this balance as AI tools become more deeply connected with everyday financial operations.
AI Demand Pricing: Walmart And Coca-Cola’s Powerful Retail Shift
