September 1, 2021 | Michael Parcelli
Michael Parcelli, senior vice president and head of global solutioning and process consulting at Xceedance, recently spoke with Insurance Post about the role artificial intelligence (AI) plays in the insurance industry today and how that role may evolve. Below are a few highlights of that discussion.
The increased use of diverse data and AI can strengthen decision-making, enable hyper-personalization, and make ‘real-time’ insurance a reality. However, the unknown risks of AI must be identified and mitigated before the technology is embedded across critical business workflows. The potential benefits of AI are well known, but the technology has its critics. The most common argument against the widespread rollout across the insurance value chain is that human errors can be controlled and mitigated, however that is not universally true. AI can have vast, and costly, consequences if not tested and managed throughout the implementation. The risk is the reliance on biased or bad data.
AI is still in the early stages
The reality is, for all the talk about AI, robotics, and real-time insurance, the industry is still in the dawn of what tech can do, according to Michael Parcelli, senior vice-president at Xceedance. “The industry is at the first level of maturity. Today we are using AI in very objective, manual task-related processes. For example, we are ingesting data from a form, then uploading and interpreting it. We have simply automated data entry.”
“There are great advances in AI technology that will allow us to bolster decision support systems, bringing consistency and reducing variability in critical business processes, such as premium generation and claims assessment,” Parcelli continued. “But decision support is not well developed in insurance – a human coming to a conclusion based on information generated by AI is the goal. It has developed to some extent, but we are not yet at the stage of multi-variate analysis.”
In the near future, AI will continue to act as a tool for decision support in key insurance functions, such as underwriting and claims. For all the fears about where AI could lead us and cause a jobs Armageddon, it is likely humans will remain in charge and dictate whether AI behaves in ‘good’ or ‘bad’ ways. The proliferation of AI could offer the industry an opportunity to create efficiencies, enhance underwriting outcomes, and deliver superior policyholder experience. “This is a chance to reset and take a different approach, to build frictionless processes, reduce complexity, and boost business agility. I hope as an industry we carefully measure our approach for a successful implementation,” concludes Parcelli.
To read the full article, Intelligence: Responsible AI, please visit the Insurance Post website (registration/subscription required).