Standard Life is testing new AI tools and automation to improve customer experience and increase people's engagement with their pensions online, according to Andy Young, head of digital and user experience at Phoenix Group,
Over the past year, Standard Life, part of Phoenix Group, has been working on developing tools using generative AI.
One such tool is to help its team navigate the compliance process to help achieve quicker publication times.
According to Young, the existing manual system currently sees as many as one in five draft items having to be reworked because of elements such as missing information required under COBS rules and consumer duty.
It is understood that reviewing content can sometimes take several days, checking information and requesting changes.
The AI tool however could help the teams check draft materials against the rules outlined in the FCA’s playbook, and make recommendations to the author before they submit it for human review.
According to Young, this would ensure the process is made easier - and faster - while being fully compliant.
This is being road-tested with the team internally, and Young said he was "really excited by it on a personal level".
Similarly, the Standard Life team is testing a proof-of-concept chatbot that can help answer questions that are most commonly asked by customers.
It is understood this would build on the existing virtual assistant pilots to enable more nuanced and tailored responses to questions and to help pre-empt demands.
Typically, internal testing takes several months, at least until it meets the OKR criteria, but also using scale testing and user testing.
Young added: "But we'll wind down experiments that aren’t delivering benefits to the customer and use feedback to give us a higher chance of developing features customers love."
Human in the loop
Ultimately Young said this was not about simply developing tools and features for the sake of being different.
"Everything comes down to three things: the customer experience - such as personalising the journey; outcomes, such as conduct risk measuring or tracking for vulnerability; and the operational capacity to make things better for the user."
According to him, this is why there will always initially be a "human in the loop" in these trials, when it comes to using AI, because these developments are being created for a human end user. The aim is to build confidence, so this intervention may be reduced.
He said: "I see innovation as either disruptive or iterative.
"While there will be jumps in technological innovation where there has been a significant disruption, along the way there will be constant points of innovation.
"For example, we can continue to innovate every day so that those 432 releases that we do each year could become 600 or 800, constantly improving the app for the end user."
These could involve more videos, podcasts, updates or content to help inform and engage users on their savings journey.
