Brokers at risk from artificial intelligence – Axa IM

Brokers will need an absolute focus on offering customers added value as their role gets squeezed by artificial intelligence (AI) and advanced technology, says digital expert Jerome Beranger, speaking at the Reinsurance Rendez Vous in Monte Carlo.

Beranger, one of Europe’s top advisers on digital change and co-founder of consultancy Algorithm Data Ethics Label (ADEL), predicts huge change is coming to society.

Algorithmically-run AI will do many jobs once taken up by humans, including roles across financial services, technology evangelists predict.

“All this is going to reduce the interactions between humans and intermediaries. We need to enact the right regulations because all this will deliver a significant impact on society,” Beranger told an Axa Investment Managers briefing on digitalisation and reinsurance in Monte Carlo.

Brokers have the potential to thrive if they focus on having a “better understanding of clients”, so they can offer added value.

Blockchain will be another technology having a big effect on brokers and society at large, according to Beranger.

Blockchain creates an automated ledger for contracts that is purportedly bullet-proof from fraud and tampering.

Insurers and brokers are already beginning to test blockchain as a platform to put commercial insurance product documentation online, such as contractor’s insurance.

“It will be possible to develop transparency and trust because of its greater security,” he added.

Despite the promise of technology, Beranger pointed to a range of risks around the design of AI.

UK insurers are using the latest technology for underwriting and pricing, leading to questions over ethics.

For example, an investigation by UK tabloid The Sun revealed that Hotmail email users faced Admiral motor premiums up to £30 higher than Gmail users. Admiral has admitted it selects risks based around domain use.

Asked whether these types of ethical risks could plague insurers, Beranger said artificial intelligence was run by algorithms created by people.

It was important to question their ethnicity, diversity and outlook to ensure the artificial intelligence worked within the right ethical boundaries.

“One must create the right framework for them to integrate the right recommendations. The design of an algorithm must be based on diversity and culture,” he said.