Artificial intelligence is poised to offer insurers valuable growth opportunities for revenue and employment, according to Accenture’s new research,
The study, which surveyed 1,200 CEOs and top executives working with AI, as well as 14,000 workers spanning four generations and representing all skill levels, found that insurers that invest in AI and human-machine collaboration could, over the next five years, boost their revenue 17 percent and their employment 7 percent. In addition, the research showed that most insurance workers are enthusiastic about the potential of AI, with two-thirds believing it will create opportunities in their work (such as underwriters using AI to generate more automated data aggregation about exposure, or for claim handlers to glean data-driven insight at the right time in a claim file).
However, insurance companies are hardly leaping towards training their workforce to implement and collaborate with AI, says Jim Bramblet, managing director at Accenture. The study found that insurance executives believe only one in four employees are ready to work with AI, and almost half cite the growing skills gap as the number-one factor influencing their workforce strategy. Yet, only a paltry 4 percent have current plans to train their workforce to implement AI.
This response was surprising given the potential benefits of AI in the workplace, he says, and speculated that issues related to prioritization and how to deliver training could be at play. “Part of it could also be trying to understand what skills need to be trained and how to start thinking about a curriculum,” he explains.
But it’s essential that insurers consider changing that investment profile, sooner rather than later, he adds, or companies risk missing an essential growth opportunity. “If you look at the macro components tied to talent in insurance, there is an aging and declining workforce,” he explains, pointing out that the industry needs to become a more attractive place for new entrants in the marketplace. “Insurance remains an industry with a strong message of being here for the greater good and is in a unique spot to be able to connect that message with the fact that now there can be valuable intelligence to serve up insights at the appropriate time to get the best possible results,” he says.
Bramblet emphasizes that investing in true AI (which Accenture refers to as “applied intelligence,” or using AI tools in combination with human ingenuity across the enterprise) differs from the investments organizations have already made in robotics process automation (RPA). “Investments in RPA left our clients wanting in terms of benefits they harvest,” he says. “Automation components are great for existing tasks but companies haven’t yet used AI to add real insights in the context of allowing the workforce to make better decisions that lead to better outcomes.”
Succeeding with AI, the study noted, will require dedicated leadership and a cross-enterprise strategy with long-term budgeting, as well as a workforce significantly different than that which is found in most insurers today.
“Many insurers are already going through agile transformation exercises that involve constructing teams and re-thinking cross-discipline skills,” Bramblet points out. “I believe the training required to get the workforce comfortable with AI fits nicely into the paradigm that companies are already investing in.” The skills required to bridge the ‘missing middle’ between humans and machines, he says, are often around softer skills such as communication, driving judgment based on data- driven insights and research-led design.
The Accenture study found that to prepare the insurance workforce for this stage, executives must reimagine work to better understand how machines and people can collaborate; pivot their employees to areas that create new forms of value; and teach new skills that will enable employees to work effectively alongside intelligent machines. “Insurance companies and associated HR organizations must take a real close look at nature of work,” says Bramblet.
