AI Stars in SLG Financial Data Extraction


The Government Finance Officers Association (GFOA) and Rutgers University have embarked on a joint project to use artificial intelligence (AI) technologies to extract key financial data from the financial reports of local governments.

The project aims to enhance the speed and accuracy with which financial data can be gathered and used to support decisions. The effort is using advanced natural language processing algorithms and machine learning techniques to extract specific financial data from local government financial reports, including revenue, expenditures, budget variances, and debt levels.

Ultimately, utilizing an AI-powered data extraction process may provide an efficient and reliable means of accessing vital financial information, reducing the need for manual intervention, and reducing the risk of human error. As it stands now, accessing financial information from local government reports is a labor-intensive and time-consuming task.

GFOA and Rutgers University want to apply AI systems to test the possibility of enhancing the efficiency and accuracy with which financial data can be gathered and used to support decision-making.

The project is currently underway, and GFOA and Rutgers expect the first results from the project in the early fall of this year.

“By harnessing the power of AI, we aim to transform the way local governments access and utilize financial data,” Chris Morrill, Executive Director of GFOA, said in a statement. “This joint effort will streamline processes, enhance accuracy, and ultimately empower finance professionals to provide better information to elected officials and the public.”

“Our AI research capabilities, combined with the expertise of GFOA, present an exciting opportunity to leverage cutting-edge technology for the betterment of government financial management,” said Miklos A. Vasarhelyi, a professor of accounting information systems at Rutgers.

“We look forward to developing innovative solutions that address the challenges faced by local governments and contribute to their financial success,” he said.

The team will work with about ten county governments to test whether data can be extracted, and then will trial the process of integrating information from the county into a large language model. The work will also focus on extracting a small number of the most critical pieces of information, including quantitative and qualitative information. The project will also explore the integration of non-financial and financial information.