December 14, 2016

AI, Blockchain, Machine Learning and the Future of FinTech


The future of financial services is both promising and exciting, especially when we look at innovations emerging in payments and commerce. We’ve seen startups like Stripe integrate more robust features, such as instant payout and fraud protection, to their platform to enhance their capabilities. Earlier this year, our portfolio company SimplyTapp launched Gane, a consumer app offering immediate access to funds, Tap and Pay POS functionality, and the ability to automatically collect and apply discount offers in a single mobile experience. As we continue to follow this evolving sector, I’ve identified a few major themes to keep an eye on:

AI is part of Fintech’s future: Large and small companies alike are leveraging artificial intelligence and natural language process to give their users better access to financial services (Click to Tweet). Larger companies such as Bank of America are looking at using machine learning to help users better understand and access their account information (for example, to automate financial goal-setting or query a certain transaction), resulting in a better customer experience.

On the smaller side, startups such as Clarity Money are seeking to generate product and service recommendations based on spending patterns and credit usage, to point out services that are ill suited to users or better recommend the type of product that is best suited for specific transactions.

Connected devices will drive the next wave of payments: The era of device-based commerce is rapidly emerging. Payments from connected devices such as watches and jewelry arguably herald the emergence of an era of smart devices that can store payment credentials as well as pay for negotiated transactions. True device-to-device payment interactions still have to overcome barriers such as the ability to securely embed payment credentials and seamless authentication between devices. In the interim, connected devices will continue to authenticate themselves via the cloud before negotiating payments and transactions between themselves.

Users need greater incentives to use mobile payment in stores: Point-of-sale mobile payments still appear to lack that set of killer use cases to spur adoption. Although businesses are investing in EMV acceptance terminals at the point of sale, the experience lacks value drivers to encourage users to transact with their mobile devices. According to Accenture, loyalty, integration, ticketing, special offers, ID and checkout could play a major role in accelerating the adoption of mobile payments in retail stores. If that integrated value proposition can be delivered, then there is a chance that Accenture’s prediction of a 50 percent increase in North American mobile wallet users by 2020 may materialize (Click to Tweet).

Blockchain’s promise of effective, low-cost transaction verification is catching on: Blockchain technology’s role in providing a distributed ledger as a system of record is attracting broad interest in the payments and commerce ecosystem. The technology threatens to disintermediate the role of fee for service 3rd party providers of transaction authority.Use cases gaining traction include transaction settlement and clearing, and syndicated lending. Consortiums, such as R3, that bring together large banking institutions in support of blockchain will help gather momentum and accelerate acceptance of the technology.

Machine learning could improve payment security: Payment security approaches will need to evolve with the digitization of payments and an era of payments from a broad range of connected devices (Click to Tweet). Devices may have increased risk of exposure to advanced attacks, which are not signature-based in the hope of initiating fraudulent transactions from these devices, many of which have not been designed with the idea of payments security in mind.

While designing high-end payments security for devices such as your car or appliances in your connected home will be a gradual evolution, machine learning offers the promise of establishing a baseline of normal device behavior and inferring when such behavior has been compromised via the detection of unusual behaviors or transaction activity.  


Tags: fintech , blockchain , AI , machine learning , Suresh Madhavan