Citibank Secured Patent for AI-Drafted Financial Reports
The newly patented system utilizes large language models to automate suspicious activity reporting for the U.S. Treasury.
Updated on Oct. 14, 2026 in Artificial Intelligence
Live Poll
Do you trust artificial intelligence to correctly identify criminal activity in financial reporting?
On September 22, 2026, the U.S. Patent and Trademark Office issued a patent to Citibank for an AI system designed to draft suspicious activity reports. The technology aims to address data processing inefficiencies by automating the generation of reports for the U.S. Treasury Department.
Why it matters
Financial institutions face significant manual labor burdens in compliance reporting, where internal data suggests that as few as 4% of filed reports actually require law enforcement follow-up. This system seeks to minimize human error and time expenditures by automating the report drafting process.
The system processes transaction data from customer devices and internal databases before a prompt generator translates the information for a large language model. The patent application process for this technology concluded after 666 days.
The players
Citibank
A global banking institution that develops financial services and digital compliance infrastructure.
U.S. Patent and Trademark Office
The federal agency responsible for granting patents and registering trademarks in the United States.
The details
The system acts by ingesting raw transactional data and converting it into instructions that a large language model — a type of artificial intelligence trained on vast amounts of text to predict and generate human-like content — can utilize. This mechanism produces a machine-readable report for the U.S. Treasury Department. The process is intended to streamline compliance workflows by reducing the manual overhead previously required to interpret and document financial transactions.
Timeline
November 2024: Citibank filed the initial patent application.
September 22, 2026: The USPTO officially issued the patent.
The Tech Race
This patent follows the ongoing industry trend to automate Bank Secrecy Act compliance requirements through artificial intelligence. It highlights the growing competition among financial institutions to replace legacy manual reporting systems with machine-readable automated outputs.
While this system targets institutional compliance, its deployment potentially increases the speed and accuracy of financial monitoring for the U.S. Treasury. There is no announced date for when this specific model will be operational in active banking workflows.
The takeaway
Financial institutions are increasingly looking to generative AI to filter high volumes of data for regulatory compliance. Watch for whether this patent translates into a deployed software platform for Citibank's internal reporting teams in the coming year.
Further reading
For more on how generative models are reshaping compliance, visit the Artificial Intelligence section.
Source note: This article includes information reported by Dallas Innovates.
Live Poll
Do you trust artificial intelligence to correctly identify criminal activity in financial reporting?
