BNP Paribas is testing the extent to which AI can be integrated into the day-to-day workings of investment banking. According to financial newsThe bank has rolled out an internal tool called the IB Portal, designed to help bankers build customer pitches faster and with fewer iterations.
Pitch preparation is at the heart of investment banking. The team compiles market views, trading history, and customized narratives under tight timelines. Much of that effort repeats work that already exists elsewhere in the organization. Slides, graphs, and antecedent analyzes are often rebuilt from scratch, even if similar materials have been previously used by another team or office.
The IB Portal aims to reduce that waste. The system searches BNP Paribas’ past pitch decks and uses what the bank calls “smart prompts” to display relevant slides, analysis and supporting content for new assignments.
George Holst, head of BNP Paribas’ corporate clients group, said the tool works like an AI-powered search engine, helping bankers find what’s important before pitching or meeting with clients. He says it can cut research time by days, giving his team more time to focus on strategy and client decisions.
Use cases are important because they place AI inside rather than around real-world, constrained workflows. A pitch deck is not a typical document. These reflect internal perspectives, client-specific details, and regulatory requirements. The utility of AI tools in this situation depends more on structure than conversational flair. This includes deciding which materials will be searchable, setting clear access controls by region or business line, and defining how captured content will be transitioned from internal drafts to client-ready output.
In practice, it also means traceability. Bankers need to see where the information comes from, and anything generated by the system needs to be reviewed by a human before it leaves the company. Without these checks, the risk of mistakes and inappropriate disclosures increases rapidly.
BNP Paribas builds AI tools on internal platform
The portal also fits into BNP Paribas’ broader internal build-out. In June 2025, the bank outlined its “LLM as a Service” platform, which aims to provide business units with shared access to large-scale language models located on the group’s own infrastructure.
The platform is operated by an in-house IT team and hosted in BNP Paribas data centers with dedicated GPU capacity. The bank said it supports a combination of models, including open source options and systems from Mistral AI, and plans to add models trained on internal data. Possible use cases include internal assistants, document creation, information retrieval, and more.
Other major banks are taking a similar approach. JPMorgan Chase noted increased use of its internal “LLM suite,” which allows staff to access models in a controlled environment. Reuters reported on Goldman Sachs’ investment in AI engineering and the rollout of its own GS AI Assistant.
UBS discussed internal M&A “co-pilots” used for idea generation. Alongside these internal efforts, specialized tools like Rogo are gaining traction at companies like Nomura and Moelis, pointing to the demand for financial-specific AI tools.
For BNP Paribas, the real test will be whether the IB portal becomes part of its daily operations rather than a one-off experiment. The potential benefits are straightforward. You spend less time searching, there are fewer duplicate decks, and your organization’s knowledge is more reused. The risks are equally well known. Illusory data, obscure sources of information, and accidental disclosure of sensitive information all have real implications for banking operations.
The most stable deployments have severely limited AI. This typically means pinning output to approved internal content, applying role-based access controls, logging tool usage, and requiring human approval before anything reaches the client.
When the IB Portal operates within these boundaries, it provides a practical view of how enterprise AI is being shaped. Not as a source of instant answers, but as a faster and more secure way to navigate what your organization already knows.
(Photo provided by Enrico Frascati)
SEE ALSO: CEOs are still betting big on AI: 2026 strategy and return on investment
Want to learn more about AI and big data from industry leaders? Check out the AI & Big Data Expo in Amsterdam, California, and London. This comprehensive event is part of TechEx and co-located with other major technology events. Click here for more information.
AI News is brought to you by TechForge Media. Learn about other upcoming enterprise technology events and webinars.
