With an increasing share of insurance executives who view obstetric intelligence as a tool to simplify jobs and improve jobs, such as fraud detection, Deloitte expects that artificial intelligence technologies can save the property/casualty insurance/industry tens of billions of dollars in the next few years.
Deloitte wrote in a report that by implementing the technologies driven by artificial intelligence through the life cycle of claims and integrating the actual time of multiple methods, P/C insurance companies can reduce fraudulent claims and save between 80 billion dollars and 160 billion dollars by 2032. Fraud detection systems, according to the report.
“Multiple technologies such as automatic business rules, integrated artificial intelligence and machine learning methods, mining texts, abnormal detection, and network link analysis can record millions of claims in real time,” said Deloitte in the report. “Incorporation of data from various methods, such as text, photos, sound, and video, can help identify patterns and anomalies and enhance the investigation process by reducing the number of false positives, increasing the rates of detection of fraudulent claims, and providing costs related to investigations in fraud.”
Deloitte data indicates that soft fraud, which includes an enlarged legitimate claim and represents 60 % of all accidents, has a discovery rate ranging between 20 % and 40 %, according to Deloitte data. Distical fraud, which was identified by taking deliberate measures to create false claims, stimulated 40 % of the claims of claims. It has a detection rate between 40 % and 80 %. In a survey in June 2024 conducted by DELOITTE, 35 % of the insurance executives chose to discover fraud as one of the five most important areas for developing or implementing GEN AI applications next year.
Kedar Kamalapurkar, the administrative director and leader of Deloitte Consulting LLP, said that artificial intelligence technology can address difficult and sophisticated fraud and help prevent claims.
Difficult fraud
Kamalapurkar explained that with the progress of technology, it can become a difficult fraud. Obstetric artificial intelligence can create difficult-to-discover images. Either by repeating damage or causing old damage that looks new, fake images cannot be registered with artificial intelligence systems from sellers who create estimates based on images.
Adding digital fingerprints to accepted images is one of the ways that Kamalapurkar has seen that the demand for technology begins to catch up. The fingerprints work as the DNA lines that alert companies when providing repeated images. Kamalapurkar also witnessed that the sellers share fingerprints through the client’s bases to expand the examination.
He said: “These companies have been present for two years, but I believe that the ability to integrate them into the process without adding a lot of additional time – and this allowed him to become more efficient.”
He did not remember seeing the artificial intelligence model he could determine whether the image is real or fake across more than one sub -vehicle for six or eight months ago. Since then, models have become more effective in discovering Deepfakes.
“With more virtual you get, it increases the possibility of facing some of this.” “Our ability to discover it. Artificial intelligence will be very important because it can find almost at the pixel level … a variation in a picture. Or discover that the whole image is created by artificial intelligence.”
Soft fraud
Proof of soft fraud is more difficult, but artificial intelligence provides some hope for the claims.
Kamalabork said that artificial intelligence models have the ability to collect different parts of information that previously required more cost and effort. This data can provide visions that previously require more work and higher costs, such as taking advantage of advanced sensors and video cameras to determine what happened in an accident, which was wrong and how the effect was related to causing injuries. He said that to do this, companies link the effects of cars from vehicles, estimates and damage to medical information.
“There are models on the market more objective, supported by medical science, proving what is likely to have happened in exchange for not.” “It still requires this thinking,” if you go to a jury, does anyone care? “This is what the debate is.”
Regardless of any discussions on the extent of the value of technology that will eventually prove that it is in the court environment, these data visions provide a deeper perspective in the claim. Even if they do not stop fraud, Kamalburkar sees the implementation of this type of technology as a deterrent. He said that the implementation of anti -fraud technology raises the entry barrier for hard and soft fraud, and the fraudsters are sent to other goals.
protection
Kamalapurka also believes that Amnesty International can prevent claims by coordinating personal information and alerts to documents holders: Combining historical claim data from an insurance company and a sensor from inside the car, which can demand summons or service discussions on an individual basis. Website information can encourage drivers to stand in many accidents that are home to fewer accidents.
“I think my wide idea of artificial intelligence in this case is that a person will be better than a person alone or artificial intelligence alone,” Kamalburkar said. “Because you need a context and experience will both be.”
