NewsBite

Advertisement

This was published 1 year ago

ASIC wants to ‘intelligently automate’ everything. Experts aren’t sure that will work

By Charlotte Grieve

The corporate watchdog will increase the use of automation and artificial intelligence (AI) across its operations, raising concerns among experts that the required technology remains under-developed and could allow serious allegations of white-collar crime to evade investigation.

The Australian Securities and Investments Commission (ASIC) has listed “intelligently automate everything we do” as one of five principles of its internal digital strategy, according to documents obtained by this masthead under freedom of information laws.

The use of automation within ASIC garnered controversy earlier this year after liquidators claimed reports of serious corporate misconduct were dismissed in as little as 38 seconds, allowing thousands of company directors to avoid scrutiny over five years.

The documents show that ASIC is both prioritising the use of controversial computing processes at the same time as failing to invest in technology at the same scale as its global peers.

The moves come amid a Senate inquiry into ASIC’s effectiveness. It follows the inaugural review by the Financial Regulator Assessment Authority (FRAA) last year that found ASIC had under-invested in technology and had poor digital literacy among its leadership.

In response to the authority’s report, ASIC held a 23-week campaign focused on “building the digital literacy of senior leaders” by sharing case studies, articles and information about topics such as “digital mindset, innovation, automation and artificial intelligence”, according to documents.

ASIC chairman Joe Longo.

ASIC chairman Joe Longo.Credit: Alex Ellinghausen

The regulator also launched an internal campaign to “identify automation opportunities”, which produced 134 “automation ideas” which were assessed and costed. A staff presentation from last October showed ASIC plans to use automation in “high-volume and error-prone manual processes”, but did not specify which functions this included.

A spokesman for ASIC said the regulator’s use of AI was aligned with Australian AI Ethics Principles developed by the Department of Industry, Science and Resources.

Advertisement

“There are many ideas under consideration, in various stages of evaluation, and we will have much more to say on this in due course,” the spokesman said.

Helen Bird, who sits on ASIC’s corporate governance panel and is a senior law lecturer at Swinburne University, warned the technology behind automation and AI was not ready for wide deployment within the regulator.

“It’s not very sophisticated. I don’t think it’s at a sufficient quality that we know it’s going provide good insights for ASIC,” Bird said. “It will strengthen over time, and be more accurate as a consequence. But what does that mean for the moment? It means the regulator must act with caution and not use AI exclusively until they’re confident it’s thorough. If not, cases might not get investigated that warranted proper investigation.”

Bird said it was inevitable that AI would eventually be used across all regulators because of the significant cost-saving opportunities from removing time-consuming manual tasks and reducing or redeploying staff.

“Do I think it will lead to an uptick of enforcement? That is hard to predict,” she said. “There’s no single answer to define what is a good or bad regulator.”

University of Wollongong associate professor and corporate regulation expert Dr Andy Schmulow was more critical, and said ASIC’s focus on automation was a “distraction” from its longstanding failure to act on reports of misconduct.

Loading

“This is part and parcel of the moral bankruptcy that is now ASIC ... Instead of doing their job, they are subcontracting their work and analysis to a machine,” he said.

“The problem with ASIC is not some confusing mystery – its problem is it doesn’t enforce the law. There are so many instances where ASIC has been told, ‘you don’t enforce the law’. I feel like a stuffed gramophone repeating it.”

The FRAA is an independent authority, established after the 2018 royal commission into banking and financial services, to scrutinise the effectiveness and capability of financial regulators, ASIC and the Australian Prudential Regulation Authority (APRA).

It found only 10 per cent of ASIC’s annual spending went towards technology, compared to 21 per cent for the UK’s financial regulator and 17 per cent for the US Securities and Exchange Commission, the documents state.

The report was released last year and called for a “substantial uplift” in ASIC’s technology investment and “material cultural change” to drive its digital transformation.

The FOI documents show that in March, ASIC chair Joseph Longo and deputy chair Sarah Court met authority panel members Nicholas Moore, Craig Drummond and Fiona Crosbie, as well as Treasury secretary Tim Baird, to discuss progress on the authority’s recommendations.

Talking points prepared for Longo for the meeting emphasise existing work and point to “some additional initiatives” to address the FRAA recommendations, including staff training on climate change disclosures, sustainable finance and cryptocurrency investigations.

“Overall, we are on track and are in a good position to demonstrate progress on all the relevant initiatives across the FRAA’s four recommendations. However, it will take time to complete many of our initiatives and to see more concrete outcomes from their implementation,” the document states.

Loading

Earlier this year, Longo warned traders and investment banks against rushing to use artificial intelligence technology. He said inappropriate use of AI could create “unintended consequences” if adopted without proper controls and governance.

Liberal senator Andrew Bragg, who chairs the Senate committee examining the regulator, said ASIC’s effectiveness was at “rock bottom” and welcomed any effort to improve operations, including automation and AI.

“Their record on law enforcement is absolutely appalling. Crime goes undetected because they don’t respond to warnings,” Bragg said. “If it’s [automation] being used to improve their own systems, it’s worth trying. Things couldn’t be worse.”

The Business Briefing newsletter delivers major stories, exclusive coverage and expert opinion. Sign up to get it every weekday morning.

Most Viewed in National

Loading

Original URL: https://www.theage.com.au/national/asic-wants-to-intelligently-automate-everything-experts-aren-t-sure-that-will-work-20230827-p5dzpl.html