The Australian Securities and Investments Commission will begin exploring the application of natural language processing to crack down on dodgy behaviour in Australia's corporate and financial services sector.
The corporate watchdog plans to conduct one or more trials of NLP technology to understand how it can be used to improve decision making and efficiency in regulatory practices that are currently conducted manually.
“With the rise of artificial intelligence and machine learning around the world across all industries (especially in the financial services sector), understanding how such technologies can be adopted by regulators is important,” the agency said.
It is appealing to the market for four-month pilots of NLP that address one or more problem statements posed by the agency, as a precursor for further investment.
“The pilot(s) would need to be conducted such that ASIC staff will be able to understand how the technology works and evaluate its potential," it said.
“An ancillary objective of the trial includes assisting ASIC to develop in-house skills and expertise in the area.”
The pilots would need to show how certain existing manual practices, like the reviewing of product promotions in the financial and credit sectors for misleading claims, could be improved through NLP.
“As social media advertising is often targeted to the individual based on their browsing history, ASIC is only getting an exposure to a small fraction of the promotions,” the agency said.
“This means that there is a potential gap that ASIC is not reviewing.”
NLP could also be used to identity poor practices during insurance phone sales; ASIC says the current manual approach is “costly and time consuming” and limits the number of calls that can be reviewed.
Pilots that assist ASIC to become more proactive in assessing managed fund product disclosure statements (PDS) and poor financial advice will also be considered.
“ASIC would like to be in a position to be more proactive in reviewing disclosures in PDS to identify any mis-selling of product, better inform ASIC policy development and risk management activities and reduce the potential for adverse consumer impact," the agency said.
It is planning to begin pilots in May.