• Finextra reports that Deutsche Bank has implemented an AI platform across its capital markets and investment banking teams in Europe and North America.
• IT Brief UK, in a report dated 14 October 2026, says the platform is Farsight's and runs on-premise.
• IT Brief UK lists the target documents as pitch books, confidential information memoranda, internal memos and financial models.
• The reports reviewed contain no usage, cost or performance figures, so productivity claims cannot yet be tested.
• One bank's deployment is a disclosed architecture choice, not evidence of an industry-wide pattern.

Deutsche Bank has put an AI platform from the vendor Farsight inside its own technology environment for investment banking and capital markets teams in Europe and North America. What the reports establish is architecture and scope. They give no user counts, costs or performance metrics, so this is a disclosed deployment choice, not proof that the tool pays for itself.

What was deployed

IT Brief UK, in a report dated 14 October 2026, says Deutsche Bank is running Farsight's platform on-premise for investment banking and capital markets teams in Europe and North America. Finextra's summary describes the same scope: an AI platform implemented across the bank's capital markets and investment banking teams in both regions. IT Brief UK adds that the system offers a chat interface that draws on context and user feedback.

Why on-premise is the detail to read

On-premise means the software runs inside the bank's own technology environment rather than on a vendor's shared cloud. For teams handling confidential deal material, that is a data-control choice, though neither report states the bank's reasons, and this piece does not infer them. What the reports do establish is the deployment model, and that is the fact other banks can compare against their own data-handling policies when they assess similar tools.

The work it targets

IT Brief UK lists the target documents as pitch books, confidential information memoranda, internal memos and financial models, handled through tools such as PowerPoint, Excel, Word and email. That is document production, the standards-heavy layer of deal work. George Tierney, a Deutsche Bank managing director and global COO for M&A, is quoted by IT Brief UK on the bank's standards for client-facing material, and the reporting presents the platform as a way to apply AI to those workflows.

What remains unproven

The reports reviewed for this piece contain no adoption rate, hours saved, error rate, contract value or running cost, so any productivity claim is intent rather than measured result. Deployment scope is stated; outcomes are not. A single deployment also cannot show an industry pattern, and this piece draws no such conclusion. Treat the rollout as a disclosed architecture choice at one bank until the bank or the vendor reports measured results.

What to watch next

Three disclosures would turn this from an announcement into evidence: usage or review-time metrics from Deutsche Bank, a second named Farsight client, and any statement on how compliance review treats AI-assisted client material. Banks evaluating comparable tools can use the deployment model as a procurement question, asking vendors whether the software can run inside their own environment and what that does to cost, without assuming one answer suits every institution.