AI & Collaboration to Combat Insurance Fraud’s Future

Earlier this month in Johannesburg, the Insurance Crime Bureau [ICB], with partners like SAS, hosted their annual conference themed “Breaking Barriers & Building Bridges for the Future”. Over 200 industry professionals, including insurance investigators and government and law enforcement agencies, attended to network and collaborate on combating the rapidly evolving fraud landscape.

Nick Feast, Principal Business Solutions Manager, Risk, Fraud & Compliance at SAS, emphasized the importance of collaboration among insurers. He noted that working alone, an insurer’s fraud team might lack historical context or miss connections that could reveal a larger criminal network. However, a consortium of industry players can piece together comprehensive claim histories and identify patterns that an individual insurer might overlook. Consortia also gather new data through anonymous tip lines to enhance fraud detection efforts.

Feast pointed out that while some insurers have advanced analytical capabilities, many still rely on manual investigations and siloed approaches. He acknowledged the progress made with AI-enabled data analytics in fighting fraud but noted that not all investigation units use AI analytics at scale yet.

Feast highlighted the value of in-house data scientists and skilled investigators, stating that technology should augment their work rather than replace them. He stressed the importance of combining technology with human expertise to combat fraud effectively.

The conference also discussed the impact of generative AI [GenAI] on fraud, noting how easily synthetic identities and forged documents can be created, posing significant threats to the industry. Feast mentioned that insurers must adopt sophisticated capabilities to detect such fraud and integrate large language models [LLMs] with decisioning platforms to enhance customer experiences through chatbots.

SAS is focusing on GenAI initiatives, including synthetic data generation to aid in model building while protecting data integrity. They are also developing co-pilots to streamline daily tasks in the analytics lifecycle and business-level operations.

Feast concluded by reminding insurers that AI requires good data to be effective and that analytics must be trustworthy, transparent, and continuously monitored.

Related Posts