BNP Paribas has integrated artificial intelligence into the daily tasks of its investment banking division by launching the IB Portal, an internal platform aimed at accelerating and simplifying the preparation of client pitches. As financial institutions seek efficient solutions amid growing data demands, tools like IB Portal reflect the sector’s evolving approach to technology. Unlike general-purpose AI apps, this system targets the repetitive challenges faced by bankers, offering practical support for teams working against tight deadlines. Using advanced search capabilities and so-called “smart prompts”, the portal seeks to promote better use of existing resources and institutional knowledge.
Other global banks have publicized similar internal advancements in recent months. JPMorganChase, Goldman Sachs, and UBS have each developed proprietary AI platforms serving various functions from document generation to idea assistance. However, earlier announcements often emphasized broad potential rather than reporting on integration into concrete daily routines. While the finance industry has historically approached new technology with caution due to compliance and data sensitivity, the competitive landscape has driven more rapid experimentation with sector-specific AI, as shown by the recent mention of tools like Rogo among Nomura and Moelis. The degree to which these tools are embedded into standard workflow remains a key differentiator.
How Does IB Portal Operate in Practice?
Within BNP Paribas, IB Portal is designed as a search and retrieval platform that mines the bank’s historical pitch documents and analysis, providing recommendations relevant to a new mandate. Rather than generating content from scratch, the tool locates and suggests existing materials previously prepared by other teams, which can be adapted for current needs.
“We see IB Portal functioning as an AI-powered search engine, enabling bankers to quickly identify what’s important for client meetings,”
stated George Holst, who leads the corporate clients group at the bank. According to the company, this ability to repurpose and adapt content aims to minimize inefficiency and redundant tasks.
What Safeguards Are in Place for Content Accuracy?
Ensuring traceability and approval remains central to the tool’s deployment. The system not only tracks the origins of all retrieved materials but also requires manual review and validation before any information is shared externally. This approach addresses the critical need for accuracy, as errors or inappropriate disclosures can have significant regulatory or reputational consequences. Access rights are managed by business unit and geography, controlling who can retrieve which documents or use the system’s features.
How Does IB Portal Fit Within Broader AI Adoption?
BNP Paribas’s introduction of IB Portal fits into its broader ambition to deploy large language models securely within the organisation’s own IT infrastructure. Their “LLM as a Service” platform supports both industry-standard and custom-trained AI models, giving various business units the flexibility to develop specialist assistants, drafting tools, and data retrieval services without moving sensitive information outside the company.
“This framework allows us to innovate while prioritising security and internal oversight,”
the bank explained. This progression mirrors the internal AI strategies being quietly adopted by other major banks, highlighting a move toward tailored solutions rather than off-the-shelf tools.
As financial institutions deepen investment in AI, the discussion has shifted from experimentation to everyday operations. The evolution of tools like IB Portal demonstrates how adoption is shaped by regulatory constraints and the necessity of workflow integration rather than novelty. For readers tracking technology in financial services, it is worth noting that value often stems from optimising existing resources and tightening risk controls, not from chasing disruptive change. Awareness of data governance, human validation, and access controls are essential for any organization considering similar AI deployments. Thoughtful implementation allows teams to reduce inefficiencies while upholding standards in accuracy and compliance.
