The rise of AI was presenting major risks in the finance sector, the Australian Securities and Investments Commission's executive director of enforcement and compliance, Chris Savundra, said.
They included problems around advertising, complaint handling and credit scoring, as well as the cyber threats it posed to financial systems, he told a federal parliamentary inquiry into AI on Tuesday.
Notably, Mr Savundra said it was also putting consumers at a higher risk of being scammed.
"We do see ... an increased use of AI, the use of big fakes, and more sophisticated ways in which scammers can take advantage of Australian consumers," he said.
Mr Suvundra said another risk was that Australians - particularly younger people - were increasingly relying on AI for financial advice.
He called for the federal government's new unfair trading practices scheme, which targets issues such as subscription traps, to be extended to the financial services sector.
The inquiry earlier heard from eSafety Commissioner Julie Inman Grant, who urged the government to avoid playing catch-up on the technology, as it had with past online developments.
"We've run this experiment before," she said.
"For years, regulation could catch up with social media harms after the damage was visible in children's mental health and election integrity and trust in our information ecosystem.
"AI is moving faster and its harms could be much more catastrophic."
She noted that the regulator has already seen AI weaponised to create child sexual exploitation and abuse material, noting there was a 75 per cent increase in reports between 2024 and 2025, particularly in schools.
In response, the commission began working with the National Home Safety Education Council, a group of 27 departments and organisations, assisting in the rollout of a school-based reporting system.
Enforcement action against several so-called "nudify" services had resulted in eight of the most widely used programs withdrawing access from Australia or introducing age restrictions and stricter standards.
The commission had a 96 per cent success rate at getting image-based sexual abuse material taken down.
But that action was only possible if there was a sexual component to the material, a gap Ms Inman Grant acknowledged.
"We're seeing image manipulation of teachers or principals doing and saying things they didn't do, which can be harmful and defamatory and hurtful to their reputations," the commissioner said.
"But without that sexualised component, we don't have powers to be able to remove that content."
Other content could only be removed if it met the high standard for adult cyber abuse, a power the commission had only been able to use successfully in six per cent of cases.
There was also a gap in regulation around the use of deepfakes in an election context, but Ms Inman Grant noted the issue fell outside the commission's remit.
The commission also heard from the CSIRO, which acknowledged the issues associated with bringing AI computing capacity to Australia, including the need for resource-hungry data centres.
But the national science agency said such access would greatly benefit a lot of its research.
The commission will explore how it can extend existing standards to address issues surrounding AI chatbots under proposed digital duty of care laws.
The proposals, introduced to parliament on Thursday, will put the onus on platforms - rather than users - to ensure online environments are safe.
Lifeline 13 11 14
Kids Helpline 1800 55 1800 (for people aged 5 to 25)
