’It’s important not to try and re-engineer all of the processes at once,’ says divisional director

Brokers have been told that a “test and learn” approach can be key when implementing artificial intelligence (AI).

That was according to Jo Sykes, divisional director of the UK broker channel at Markel, who said that AI can present “huge opportunities” if used in the right way.

Sykes was speaking during a panel discussion entitled What’s next for AI session? at the 2024 Biba Conference in Manchester.

A key talking point during the session were the opportunities on offer to brokers who adopt AI.

Sykes said Markel uses a ”test and learn” approach to ensure it is being used in the most effective way and that there were opportunities for brokers implementing AI in a phased approach.

“There’s huge opportunity,” Sykes said.

She continued: “From a Markel perspective, we take a test and learn approach. AI can help identify which risk is the best to go for.

“From a broker perspective, you can get AI to tell you how likely [a client] is to renew, which is incredible intelligence.

”[But] I think it’s important not to try and re-engineer all of the processes at once.”

Insurer uses

Markel has been using AI for several years, with it teaming up with insurtech Cytora back in 2021 to use its risk assessment platform.

Sykes said the biggest benefit of ”AI is time”.

“What we are aiming to do is remove admin from underwriters,” she said.

“[This] is so non-value add tasks that can be performed by AI behind the scenes.”

Paolo Cuomo, executive director of Gallagher Re, added that AI was becoming valuable for fraud detection, but warned that fraudsters were also starting to adopt the technology.

Allianz, for example, said in April 2024 that fraudsters were continuing to evolve their tactics, with it noticing scams involving AI and deep fakes.

In turn, the insurer partnered with Clearspeed to enhance its fraud detection capabilities.

“AI has been used in fraud detection for a number of years, but now the bad guys are also using it to make sophisticated fraudulent attempts,” Cuomo said.

Humans

While AI is on the rise, however, Cuomo said that humans were still needed.

To stress this, he gave the example that a policyholder could incorrectly be told by a chatbot their claim does not fit the policy terms and conditions.

He said: “For complex claims, if you are an experienced claims handler, you may be handling hundreds of claims.

”A machine can upgrade every single claim that goes through your business. Therefore, it’s not replacing the human, it’s saying there was a claim three years ago that may be relevant – that’s where it’s going to make a difference.”

Sykes added: “There’s a perception that AI is going to replace people, but it complements people if you implement it in the right way.”