AI Experts Aim to Combat Tax Loopholes Costing Billions
AI helps with taxation
A team of AI researchers and experts from John Hopkins Law School aim at creation of an AI system to reduce tax loopholes. Currently, the result of tax laws manipulation is around $500 billion.
The researchers called the system Shelter Check. It will help the regulators to examine different types of unintended tax evasion.
“We think of it as a grammar check, but for tax sheltering”, says Benjamin Van Durme, a scientist from Johns Hopkins. The purpose of the system is to consider the offered changes in the law and inform the regulators about possible issues for the tax law or unplanned side effects.
It is possible to evade tax paying legally by maneuvering through different government resolutions, be it Congress or the IRS or court decisions.
Making an AI, that can navigate through tax laws as a professional accountant is a tough task. The language of law is convoluted and the tax law has tens if not thousands of pages with data for understanding tax outcomes. Currently, there is experimentation with ChatGPT series that were tail-ended by the tax code.
The existing GPT models cant manage the basic questions about the tax law yet.
However, the researchers are optimistic, especially when utilizing GPT-4, which is OpenAI’s latest large language release.
The buzz is there for making the project work, since the team is unsettled that corporate giants may outcompete them in developing similar systems, that can be even more efficient.
AI Catalog's chief editor
Share on social networks:
