AI use in financial data analysis
The US Financial Accounting Standards Board (FASB) and Governmental Accounting Standards Board (GASB) are looking into the ways in which investors are using artificial intelligence (AI) to consume and analyse financial data.
With the increasing popularity of ChatGPT, AI is getting a lot of airtime recently. As such, FASB and GASB see that it is important to continually monitor and understand how investors are using AI.
FASB Chair Richard Jones said a key part of the board’s investor outreach “is really understanding which data is the most important, how is it processed and when does it actually influence those capital allocation decisions.”
Our perspective on the growing power of AI for financial analysis? While one day it might be possible to perfectly scrape unstructured data using artificial intelligence, the technology is still a long way off. It will always be more efficient to first report data accurately and effectively in a machine-readable format, and make the resulting data publicly available for many different types of use. AI works best on structured data. That means three things in particular.
First, the use of AI to assist companies to help get through the simpler tagging decisions and free up time to decide on those that need significantly more judgement is a trend that will only accelerate. It will both increase the ease with which tagging can be done and permit more unstructured data to be marked up than might previously have been possible or economic {Ed: More Earnings Releases then?}.
Second, for digital business reporting, consistent, high quality XBRL taxonomies are becoming more and more important in the AI era. They both drive the tools that are increasingly helping management to come to the right conclusion about the tags that they need to use and provide a comparable framework for advanced analytics, including those that rely on AI.
Companies need to appreciate that more and more, their structured disclosures are being consumed by machines on behalf of regulators, investors and a wide range of other stakeholders. With the level of innovation and investment now going into machine learning and advanced neural networks, this should be a signal to focus on quality and utility for their structured data reports.
Read more here.
