Lendi Group outlines limits for AI autonomy

20 November 2026
Share this article
Lendi Group outlines limits for AI autonomy

Lendi Group has highlighted the importance of establishing clear boundaries around AI agents, as part of new research examining how Australian businesses are building trust in the technology.

The group’s chief technology officer, Devesh Maheshwari, was among 25 Australian AI leaders who contributed to Cisco’s Permission to scale: The critical role of trust in successful AI adoption report, which found that trust, rather than technical capability, was increasingly determining how far organisations were prepared to let AI go.

“Trust determines the autonomy you are prepared to give AI. The aim is bounded autonomy, where the agent can act on your behalf but cannot go beyond the boundaries you have established,” Maheshwari said.

The report found that fewer than half of the businesses surveyed were very or extremely confident their existing foundations were ready to support more autonomous AI systems.

 
 

Privacy and security breaches were identified as the leading risks, while more than half of respondents said their infrastructure needed upgrading.

Building trust in AI

Maheshwari said security, governance and reliability needed to work together in a self-fulfilling manner as businesses increased their use of AI.

“Security protects the system, governance provides the constraints and reliability earns trust. As we move further into AI, those three things have to work together rather than be treated as separate activities,” he said, adding that business should take a gradual approach to adoption.

“Start small and develop trust. Measure the impact, expand the use of AI, and gradually increase the responsibility it is given.”

md discover

However, Maheshwari also questioned the assumption that every AI-driven task needed direct human involvement, arguing that businesses should instead identify where human judgement was essential.

“Not everything needs a human in the loop. The question is where the consequential decision sits, where a human needs to make or approve that decision, and where an agent can execute within the boundaries that have already been established,” he said.

ASIC sets sights on tech

The Australian Securities and Investments Commission (ASIC) has also put its crosshairs firmly on the responsible use of AI in finance, citing one of its 2027 priorities as: “Consumer harm arising from poorly implemented technologies, including AI and automated systems.”

Commenting on the priorities, ASIC chair Sarah Court said that consumer protection “must keep pace with the development of new technology”.

She said: “Artificial intelligence, automated decision making and chatbots may improve the delivery of financial services but do not alter a firm’s legal obligations.

“The ability to obtain human review is particularly important where customers are making a complaint, responding to a scam, pursuing an insurance claim, or seeking financial assistance.

She also said that ASIC would focus on “poorly designed or inadequately supervised systems” that “produce misleading information, make inappropriate decisions or prevent customers from accessing effective support and assistance.”

“The technology may be new. The responsibility is not.”

AI also came up in ASIC’s review of its best interest duty, noting: “While there may be significant benefits from further innovation, licensees must also ensure that their risk and governance arrangements keep pace with the adoption of AI,” the regulator stated.

“Brokers and licensees are still responsible for the quality of the credit assistance they provide, regardless of the technology they use.”

Ongoing development

Lendi Group’s contribution to the report comes as the business pushes ever further into AI adoption, as has its competitors.

The fintech was one of the first groups to announce it was to become AI-native back in 2025, and have AI extend into all areas of the business, with over 1,350 brokers across 220 stores.

Lendi Group chief executive Sebastian Watkins told Broker Daily’s sister brand, The Adviser, that while technology could not yet support the complete customer journey across every scenario, the business was setting a financial-year-end goal for full coverage.

“We can’t take a customer the whole way through across all scenarios. By the end of this financial year, our goal is to enable that, and the broker will be able to use as much or as little of that process as they want,” he said.

Lendi’s competitors have also been investing heavily.

Most recently, Australian Finance Group unveiled a new AI tool designed to significantly reduce the time mortgage brokers spend collecting, verifying and entering customer documentation.

In September 2026, rival aggregator Yellow Brick Road launched its own multi-agent AI platform, designed to speed up customer settlements and improve broker productivity.

Indeed, its CRM platform is backed by Salestrekker – also used by several aggregator groups, including Specialist Finance Group, outsource Financial, and Astute – which launched a new AI-powered tool last month to audit loan files in real time at the point of lodgement.

Meanwhile, Mortgage Choice has also unveiled a bolstered suite of AI tools and tech upgrades aimed at driving broker efficiency, reducing manual administration, and modernising consumer engagement platforms, as has Loan Market.

However, Watkins said that customers would continue to value a human relationship even as the processes behind that relationship changed.

“I think where us humans will play a very important role, I think humans will still want a human interface; they just, they’ll care less about how the work is done,” he said.

“Where brokers used to add value was two very specific things: we used to add value at the relationship level, and then we would also be experts in credit.

“Now the second part of that, there’ll be no need for that, there will be no need to have any industry or lender knowledge.

He added: “Where our brokers will be insanely important is that relationship. They will want, they will need to look after their customers, they will need to find their customers, they will want to support their customers.”

[Related: Over half of home buyers use AI for borrowing despite trust gap]

Broker DailyWant to see more stories from trusted news sources?
Make Broker Daily a preferred news source on Google.
Tags: