Flamingo Ai founder Catriona Wallace says the intermediary community is “absolutely ripe for disruption” as artificial intelligence is increasingly taken up across the insurance industry.

“We are starting to see a lot of the US insurers thinking ‘how do we go direct and how do we use AI to do that’,” she told the Australian Professional Indemnity Group conference in Sydney last Thursday.

Australian-listed Flamingo Ai, which builds software robots for insurance companies, has offices in Hartford Connecticut in the US and in Sydney.

Ms Wallace says AI is reaching across all areas of insurance, including claims and underwriting, but firms looking to introduce the technology should approach it to enhance the abilities of staff, rather than a cost-cutting exercise that involves replacing employees.

“To start there, and rip people out and put robots in would be disastrous,” she said. “It must start with the augmented approach.”

AI’s uses include amplifying existing legacy systems, sitting on top of them rather than replacing them, she told the conference.

Ms Wallace says businesses looking to introduce AI into their operations should start small and identify a business problem that has resisted other technology solutions as they explore options.

Ethical questions surrounding AI as it is introduced more widely include ensuring net benefits outweigh costs, privacy rules are not breached and providing fairness.

Potential bias within algorithms is coming under greater scrutiny as more decision making is made through machine learning, particularly within large corporations.

Guidelines could be put in place that will ensure customers are able to legally contest decisions that seem to reflect biases, Ms Wallace said.

“This will be the level that I think governments will hold companies accountable to, to try and avoid these problems,” she said.

In the insurtech sector, “stand out” firms currently using AI include Lemonade, Tractable, Cape and Progressive, she told the conference.