By David Poole, Head of Growth, MYPINPAD
Each year sees seemingly unprecedented technological advancements coming to market and 2016 was no different. It was a year defined, first and foremost, by the quest for new and dynamic user experiences. Not only did we see new augmented and virtual reality products, the Internet of Things saw dynamic new M2M technology such as smart cars and home apps.
In the background, though, security continues to be vitally important. While not as headline grabbing as the latest gadget or IoT labour saving device, security is critical in keeping the technology we use safe and reliable.
Throughout 2016, security issues have been at the heart of our fintech industry. Below, we examine four of the most pressing ones.
- Data breaches
According to the Identity Theft Resource Centre, there were 522 reported breaches by the middle of July this year, exposing over 13 million records.Worryingly, the likelihood of a material data breach involving 10,000 lost or stolen records in the next 24 months has risen to 26%.
Not only have data breaches become more frequent, but their impact has become greater.Once a data breach occurs the consequences for the affected organisation can be devastating. IBM’s eleventh annual Cost of a Data Breach Study revealed that the average consolidated total cost of a data breach for a company for 2016 is $4 million.
Data breaches will remain a part of digital business life and we can expect to see more organisations in 2017 implementing more preventive security methods, as well as new technologies being developed and implemented for this purpose.
Many companies will focus on employing stronger and multi-layered authentication, as encouraged by the Second Payments Directive (PSD2), which will also mean that even if they face some inevitable breaches, access to accounts will be nullified as the stolen partial information won’t be enough to be usable.
- CNP fraud
Last year, Aité Group reported that US credit card fraud had increased 100% from just seven years ago. The study identified point-of-sale and rising card-not-present (CNP) fraud as contributing factors; which now represents 45% of total US card fraud.
The same trend is true for the UK. Figures from Financial Fraud Action UK show that fraud has been soaring non-stop since 2006, and this figure is growing each year. In the first half of 2016 fraud had risen by 25% and card fraud, which includes card-not-present fraud, was up by 31%.
The vast majority of CNP fraud cases involve the use of card details that have been fraudulently obtained through methods such as unsolicited emails, telephone calls or digital attacks. Ecommerce is one of the main targets for fraudsters. An estimated £261.5 million of ecommerce fraud took place on UK cards in 2015, accounting for 46% of all card fraud and 66% of total remote purchase fraud.
As ecommerce continues to grow, so too will CNP fraud.Despite this, many retailers and banks are holding back on tightening their security, fearing fallout from consumers rejecting friction in their checkout experience. Yet 2017 will see merchants realising the premium that consumers place on security and their willingness to sacrifice a little convenience to gain a lot of security. The value of a brand can be enhanced by getting the security experience right.So more and more companies will continue to implement two factor authentication, as Google, Apple, Paypal, Amazon and most social media did in 2016.
- Biometrics and identity management
Apple Touch ID triggered a new biometrics security revolution, to the point where it is forecasted that by 2021, 99% of US smartphones will be biometrics-enabled.
2016 saw business explore biometrics as never before yet many are still sceptical regarding storing and encryption of biological data.
Even though it is much harder for hackers to access and use, if accessed, it is extremely valuable, since biological data can’t be changed or replaced in the event of a breach. Yet the power and simplicity of biometrics will surely see further adoption in 2017.
However biometrics on their own, like all security methods, arenot infallible. Even if sensors get stronger against fake fingerprint attacks and the technology is refined, if biometrics are to start making headway as a secure authentication technology, the technology will have to be coupled with other forms of authentication, such as password or PIN.
- Machine learning
Machine learning is a branch of artificial intelligence (AI) study that concentrates on algorithms which enable computers to “learn” without being given specific programming. Being exposed to new data enables the computer to grow, change, develop and solve problems independently of new programing.
By analysing historical transaction data, machine learning is increasingly being used to detect fraudulent payment behaviour. While it takes one person about 5 minutes to check just one transaction, a machine can check larger amounts of data in nano seconds, saving time and money and making the analysis feasible in real time to prevent an attack.
The possibilities for machine learning in the field of ID and verification are considerable. Taking, as an example, the mobile phone, we each have our own quirks whilst using our device. We will hold the device in a certain way, enter key strokes in a specific manner and have countless other characteristics that can be “learned”.
Machine learning is still in its infancy and needs to be developed in real life use cases but it could allow the inclusion of an additional layer of security to ID&V processes.If taken forward, not only would your device recognise your passcode or biometric information, it would also recognise if this information has been entered in a recognised fashion. We expect to see further developments on how AI and machine learning could be used to satisfy security needs.
Overall, in 2017 we expect to see:
- Continued growth of payment card fraud, particularly in CNP and data breaches, as well as increased investments in security and authentication measures to avoid them
- New developments in biometrics and expansion of current methods
- Broader implementation of 2-Factor and multi-layered authentication, something you have, something you know, and something you are
- Developments of machine learning applied to authentication
- Increased focus on fostering innovation and encouraging competition by regulators