How Facial Recognition Technology Can be Used in Underwriting
Facial recognition is an emerging technology that holds much promise and a wide range of current and potential applications. It is also a burgeoning market that has been predicted by Markets and Markets to be worth as much as $7 billion by 2024.
One industry that is getting ready to be impacted is insurance. The Association of British Insurers found that in 2016 there were 125,000 fraudulent insurance claims in the UK, with a total value of £1.3 billion. Insurance executives are looking for ways to use technology to reduce this expense, and facial recognition may have the answer. Furthermore, this isn’t the only way that the technology can have potential uses for insurers.
In 2020, technology is an integral aspect of any business. For those based in the UK, a reliable IT support company located in London is an indispensable resource, especially when implementing and maintaining advanced tools like facial recognition technology.
How does it work?
Developments in AI brought about by deep learning could be about to make a difference to the world. There have been considerable advances in image and facial recognition technology. Each multi-layered element in images and videos can be better understood. This means programmes can recognise faces and images in video footage to a degree of 98 percent accuracy. It can also do this at a speed 1,000 times faster than the human brain.
This biometric technology has been used in recent years to verify the identity of individuals through camera image or video capture. This may be done through two or three-dimensional capture in real time, which is checked against a database of images. The system can then grant access to an authentication system, a payment method, an application or a building. China Merchants Bank has even introduced ATMs where customers are able to withdraw money through facial recognition.
Facial recognition technology can detect emotions, and in particular suspicious behaviour. It can do this by monitoring micro expressions, eye movement, pupil dilation, gaze, speech variations and tone. It is also possible to identify seven main human emotions.
How does it apply to underwriting?
This technology can be used by insurance companies when processing claims, to assess them for their validity using scientifically verified techniques for improved accuracy.
Insurance companies can interview claimants and obtain an instant evaluation of the probability they are telling the truth. Systems developed are for the specific needs of insurance companies to assess the facial expressions of claimants for suspicious signs.
It is easy for insurance companies to do this by using their smartphones cameras to record interviews with claimants. These files can then be run through an AI-driven system that cross references them with big data. This system uses deep learning to make an instant analysis of a claimant’s responses to questions along with an assessment of their veracity.
The assessment usually comes in the form of a dashboard that offers guidance to the claims handler and makes their job easier. This is a faster and more accurate way to identify the most suspicious claimants for more thorough investigation.
Use cases
Ping An Insurance is a Chinese insurer and one of the largest financial services companies in the world. It has 184 million customers and 538 million users online. They have gradually been using more facial recognition techniques to verify the identity of their own staff and their customers.
The insurance company is also using facial recognition technology for fraud detection, by identifying micro-expressions connected with dishonesty. The President of Smart Risk Management for OneConnect, a fintech subsidiary of Ping An, announced the use of this technology. But it was in 2016 that the company announced a facial recognition-based loan product, which was the first in the world.
The company claims that it can assess risk with greater accuracy and reduce fraud. Its technology is reportedly the best available, with an accuracy rate of 99.8 percent. It also claims that online loan applications can be turned around in less than six minutes. In addition to this, OneConnect supplies facial recognition technology to other financial institutions, along with AI and blockchain technologies.
Lapetus is an insurtech startup that has developed its own facial recognition technology. This can be used to evaluate heart conditions, such as diseases related to smoking, and help insurers to set the prices of life insurance premiums. The product is called Chronos and it claims to deliver a policy in less than ten minutes.
Lapetus became partners with Legal & General America in 2018 to launch a product called SelfieQuote. Users simply need to upload a selfie, answer nine questions and wait for the expedited underwriting process to be completed. The product is mainly aimed at younger users.
It is expected that in the near future, insurers will begin to combine these technologies with data taken from fitness trackers and other IoT devices. This will do more to improve the accuracy of the underwriting process.
However, there are still some obstacles to overcome in facial recognition technology, such as Rekognition from Amazon. Studies have shown that there are some race and gender biases. It is difficult for the technology to differentiate between people of different ethnicities and genders. These challenges are likely to be overcome in the next few years, but they could pose limits on its immediate adoption.
The advent of facial recognition is inevitable, but the question that remains is when it will be widely adopted. If businesses can anticipate the developments of technology, then they will find themselves in a better position to assume future changes in the market. This will give them a competitive edge and more freedom to plan for the next new development.
—
Take a look at our page on cyber insurance for more information about what services we offer.
