
Rebuilding digital trust
How Generali, one of the world's largest insurance groups, is using AI to combat fraud
Insurance fraud is big business. A study of global claims found that around one in 30 were fraudulent, and the annual cost of insurance fraud in the US alone has been estimated at more than US$300bn.
The increasing digitisation of insurance claims has exacerbated the problem. While digital claims bring convenience for customers, they also present vulnerabilities for insurers: the widespread public availability of generative AI tools makes it easy to alter or fabricate images and documents. In a recent survey by The Fintech Times, one in five claims handlers said they suspected as many as a quarter of claims now use fake supporting documents that were created or altered using AI and digital tools.
“It’s a big challenge for insurers,” says Mario Ricco, head of group claims at Generali, one of the world’s largest insurance groups. “It’s becoming easier to create a photo showing that a car has been trapped in a flood or showing non-existent water damage in a home. All these require insurers to adapt skills of teams, workflows and tools.”
Fighting fire with fire, companies such as Generali are increasingly using AI and machine learning (ML) to combat AI-based fraud. Advanced ML algorithms can flag suspicious claims based on patterns, anomalies and inconsistencies in the submitted information. For example, metadata—the hidden information embedded in digital media—can be automatically analysed for inconsistencies and cross-referenced with other data sources, such as internal claim databases, public records, government databases and social media. Software trained on vast data sets of real and fake media can also spot anomalies in both images and documents.
The AI-based fraud detection technology that we now have can drastically reduce these kinds of digital fraud. We found one case where a single certificate had been manipulated 63 times, and we were also able to identify accident photos that were taken before the start date of a policy."
Anselmo Marmonti, SAS vice president, risk, fraud and compliance solutions
“The AI-based fraud detection technology that we now have can drastically reduce these kinds of digital fraud. We found one case where a single certificate had been manipulated 63 times, and we were also able to identify accident photos that were taken before the start date of a policy,” says Anselmo Marmonti, SAS vice president of risk, fraud and compliance solutions. SAS has partnered with a leading specialist software vendor to integrate sophisticated digital fraud detection into its platform for insurers, which is currently being piloted by a number of leading companies. “An integrated solution minimises costs for insurers and also offers the flexibility to evolve quickly as new forms of threat arise,” says Mr Marmonti. But insurers need to look beyond isolated AI tools to unlock AI's full potential. Combining detection capabilities with a real-time decisioning engine enables immediate intervention. “We see this approach driving a 20-30% increase in fraud detection rates. The reduction in claims leakage often covers implementation costs within the first year.”
Creating a connected claims ecosystem

Generali uses AI not only to verify the authenticity of claims but also in the broader area of claims management. “We are increasingly deploying AI in claims management processes to assist claims handlers to prioritize activities and make decisions faster,” says Mr Ricco. “Typically in more complicated cases there is a lot of information to assess. AI can support that.”
The combination of AI with a robust, connected claims ecosystem is now a necessity for insurers, he adds. This can be achieved using a cloud-native AI and analytics platform such as SAS.
“When we're talking about an ecosystem, it's not only about the intelligent decision-making element,” Mr Ricco explains. “The management of the workflow and the interactions with external stakeholders and sources of information are also important, including the interaction with the claimant.”
“Insurers work with thousands of entities to settle claims—body shops, medical providers or contractors—each one representing a step to make someone whole following a catastrophe,” says Franklin Manchester, global insurance strategic advisor at SAS. “Ideally, a claims ecosystem allows information to flow freely, with transparent processes, and every party in alignment.”
Ideally, a claims ecosystem allows information to flow freely, with transparent processes, and every party in alignment. Orchestrating this journey happens best when all parties have access to the data they need, when they need it, and the autonomy to act on it, with minimal manual intervention.”
Franklin Manchester, global insurance strategic advisor at SAS
When the data flow works properly, the impact is significant. Insurers who can analyse the ecosystem of provider relationships—rather than simply focusing on individual claims—are better positioned to identify anomalies in billing patterns and can quickly make huge financial savings, Mr Manchester points out.
Working with AI
Insurers considering deploying AI first need to thoroughly understand the business needs that the technology is satisfying and what solutions it can bring, particularly in claims management, says Mr Ricco.

“A lot of thought is given to the development phase and building the solution,” he adds. “But it also has to be deployed in a way that teams are confident about. If they are going to use it to assist them, they need to trust it. Teams must also be constantly trained on new patterns and have mechanisms in place to promote a proactive mitigation of any attempt of claims fraud or abuse.”
Mr Manchester adds that although teams should be given access to AI capabilities, they should only allow AI the necessary amount of authority to accomplish its specific task. “Ultimately it’s humans—and not machines—that must own the decisions.”
More from Economist Enterprise and SAS


2026 insurance trends: What comes next?
Hear insights from industry leaders on the disruptive forces shaping insurance.

The role of AI in insurance
How AI is reshaping insurance, with insights from global executives on emerging use cases.


Insurance decisions with trusted AI
Governed, explainable and trustworthy AI can unlock insurance innovation.