AIA is moving on from ‘digital transformation’ following what the insurance company deems a successful, three-year program to address its technology and data needs. Now the focus is on embedding intelligence, including generative artificial intelligence, into distribution, operations and customer services.
The company expects the adoption of genAI copilots (AI assistants) to serve as a massive productivity tool. “We need people to embrace this,” said Sue Coulter, head of group digital and analytics in Hong Kong. “It’s not about replacing employees.”
The move to embed intelligence is a big contrast to the previous years of work, which was a foundational transformation across cloud adoption, process automation, new digital applications, and increased use of data and analytics. The next three years will focus on enhancing distribution – including agents and bancassurance – and customer service.
“GenAI will have a positive impact at the heart of our business,” Coulter said.
This doesn’t meant AIA isn’t looking at many operational uses of artificial intelligence. It is rolling out a variety of pilots, from audit to assessing claims. But whereas a broad push for digital transformation is about business-wide foundations, AI can be tested and deployed for specific use cases, and boosting agent productivity offers the biggest opportunity to grow revenues.
Foundation work
Coulter has been AIA’s digital maven since 2016 but it wasn’t until 2020 that the insurer made digital transformation a strategic priority. That was both the year that the Covid pandemic exposed institutions’ digital backwardness, and when AIA appointed a new group CEO and president, Lee Yuan Siong, who took over from Ng Keng Hooi.
Under Lee, the company launched Operation Falcon as an effort focused on improving agents’ digital sales capabilities, as they struggled with the abrupt end to face-to-face customer interaction. It also included ideas around using digital tools to recruit and train agents.
Within six months, the company had rolled out capabilities such as enabling remote customer onboarding and sales. But the management also realized it needed a comprehensive digital transformation, which led to the launch of its TDA Program – for Technology, Data and Analytics – in late 2020.
This program was spearheaded by Biswa Misra, group chief tech and life operations officer, who oversees a huge portfolio that includes tech, operations, infrastructure, systems architecture, and developer teams. Coulter, who reports to Misra, is responsible for digital, analytics and AI.
Transformation didn’t come cheaply: AIA set aside a $1 billion budget to achieve a myriad of KPIs within a three-year deadline.
Mission accomplished
The program officially concluded in December 2023, and Coulter says most KPIs have been met: “TDA galvanized the organization around a set of common goals.”
While she didn’t show DigFin the results of TDA, she picked a few achievements. In three years, AIA moved about 89 percent of its compute to public cloud, and about 76 percent of its core processes for certain functions (such as claims, underwriting and customer servicing) are now fully automated.
TDA also implemented new capabilities in areas including digital sales leads, automating aspects of bancassurance (including its regional partner, Citibank), and app-store ratings: the average rose from 2.7 stars in 2020 to 4.5 stars today, Coulter says.
Getting such an expansive project to work across the entire group in a relatively short time required buy-in from the top and a framework for governance, so standards and best practices could be disseminated to all the business units.
That governance also provided a toolkit for making things happen. “It’s one thing to set a target, but then how do we ensure all of our markets achieve these goals?” Coulter related. “How do we actually do this?”
For example, it soon became clear the teams needed user-experience (UX) and user-interface (UI) capabilities. Coulter established a design ‘center of excellence’. Ditto for analytics, agile dev-ops and, most recently, AI.
Shifting benchmarks
Coulter’s background is banking – a UK native, she was chief digital officer at Credit Union Australia before joining AIA. The insurance industry is routinely said to be several years behind banks when it comes to tech. What was her take?
She accepts the premise, noting that in 2020, AIA commissioned a consultant to help with TDA that drew up a benchmark against the region’s most digitally savvy commercial banks, such as DBS and Commonwealth Bank of Australia. AIA was indeed behind, as was the rest of the insurance industry.
As the TDA Program was winding down last year, she asked for another benchmarking. The consultant now found AIA’s level of digital capability had reached that of the CBAs and DBSs of three years ago. So the firm has caught up, but of course those banks have not stood still.
However, as AIA turns its focus to AI, Coulter says matching banks is no longer the goal.
“We’re now benchmarking against companies like AirBnb, Uber and Apple, because those digital experiences are what people now expect.”
This shift is another reason why the next era of tech development is more about agents and customers.
Front-office focus
Coulter now refers to TDA as “TDA 1.0”, and the turn to AI as “TDA 2.0”. But this shorthand doesn’t capture the narrower focus of the insurer’s drive to adopt different types of AI.
“We’re still a face-to-face business,” Coulter said. “Agents are still key.”
One expression of this is the importance that AIA places on the Million Dollar Round Table, or MDRT, a US-based association of financial professionals that admits insurance agents that bring in big sums every year. AIA claims the most MDRT agents globally, a status it sought to maintain during Covid.
AIA is not unique in valuing MDRTs, as any insurer with tied agents is keen to nurture high-powered salespeople. But now digital tools are coming into focus to enrich a carrier’s agency network.
As part of TDA 1.0, AIA helped agents use and distribute content via social media, to improve customer engagement and build leads, or processes to help underwrite policies quicker.
It’s now piloting the use of AI in recruitment processes to identify people likely to become a high-selling agent. Such profiling can also identify the types of training that would benefit an agent.
When these pilots move to production, AIA will expand their use to other areas of distribution, as well as to personalizing digital engagement with policyholders, notably with regard to products and services for health and wellbeing.
“It’s exciting to see AI and the use of data moving at such a pace,” she said.
Managing AI
New tech means new risks. The firm has already established an AI Council and frameworks for managing artificial intelligence. It’s testing Microsoft copilots using internal data. The problem isn’t hallucination or errors, which occurs when language-learning models such as ChatGPT are riffing off the entire internet.
The problem is that AIA’s own documentation can be inconsistent – for example, Coulter says the pilots can dig up two documents that use different terms for the same meaning, or contradictions in procedures.
These copilots also require a degree of training. For example, using LLMs to help generate reports for in-house auditors is not just about feeding it data: auditors speak their own language, in terms of tone and phrasing, and the prompt engine must reflect this.
But once these tools deploy, they can make a huge difference in output. For example, content and marketing: the creative people can now design a single campaign, and use AI to transform that into content across 18 markets, in different languages, with locally sensitive images.
And, unlike TDA 1.0, this new strategic push doesn’t require a $1 billion budget: the capabilities already exist, the data scientists are already employed, extra products like LLM APIs can be bought off the shelf. For AIA the hope is, having invested in the unglamorous work of digital transformation, its tech is poised to reap visible rewards.
