A lot happened in the digital banking world this month, from major events in the payments worlds like PayPal acquiring iZettle for $2.2 billion only days before the target’s IPO to major regulations such as GDPR going live last week. Here are some of the important trends that Bank Innovation pulled out of the fintech chatter this month.

1.  Artificial Intelligence (AI) Implementations Advance

If blockchain was the word of 2017, artificial intelligence is the word of 2018 — or at least of May. And from the way things are going, this trend will likely carry on for the rest of the year. Applications for artificial intelligence, recently at least, have been tied to improving customer service via a bank’s digital channels. Banks and fintech startups alike have been launching chatbots and personal finance management tools, both of which use AI to interact with the customer as well to analyze data gathered in the process.

However, AI can be used for much more. In the customer-facing areas of banking, AI can be used in physical bank branches (hint: computer vision) and can also be used to reduce certain repetitive tasks for bank’s employees. For instance, French bank Credit Mutuel, recently launched a large-scale B2E centric-AI project, in which the bank is working with IBM to deploy AI to free up an employee’s time, or as Jean-Philippe Desbiolles, VP of cognitive solutions France at IBM, previously told Bank Innovation, “empower,” the employees to focus on more precise tasks. In the process of improving customer and employee experience, AI can help banks cut its costs.

In fact, according to a recent report by Autonomous, AI will help banks and FIs save as much as 22% of their costs by 2030. Taking into account all the banks across the world that amount exceeds $1 trillion, according to Autonomous. The report said that most of the cost savings will come from using AI in the frontend of banking ($490 million). However, that depends on how comfortable the consumer is with AI. The backend and middle office applications of AI are also vast and shouldn’t be ignored. These applications include using AI for KYC, authentication, compliance, data processing and underwriting. IBM Watson Financial Services, which is currently working with several banks and credit unions, is developing cognitive technology on AI initiatives for KYC, regulatory compliance as well as fraud detection and security, Watson General Manager Alistair Rennie previously told Bank Innovation.

When it comes to underwriting many fintechs are already forging ahead of traditional banks. These fintechs have been using AI to underwrite loans such as more complex SMB loans, as well as using it for alternative credit scoring methods (See trend #2).

2. Alternative Lending Back in Vogue

The underbanked market is becoming increasingly important, especially for fintechs such as Petal or Deserve, which have capitalized on banks’ neglect of a certain section of the consumer market. By using machine learning and other data points beyond FICO scoring, these fintechs can provide credit, and in other cases, even consumer loans. A lot of the alternative credit scoring comes from determining a consumer’s financial health by gathering data such as a consumer’s spending habits, income, and account balance. Machine learning, in some cases, is used to analyze behavioral patterns to predict how reliable a user is at paying their bills.

Another aspect of alternative lending is the SMB space. Here fintechs like BlueVine, Kabbage or Mirador, among others are using technology such as ML and, surprise, surprise – AI to underwrite loans. Oddly, the SMB customer is a bank’s most valuable one (more on this here), yet these banks tend to give up on SMB lending because they find the underwriting process too expensive. Small business loans are usually under $200K, but since they’re still corporate loans, they still bear heavy costs for the bank.

And that’s where fintechs stepped in to carve a niche for themselves in this expanding market.

3. Voice Banking Work Continues 

Call it Alexa anxiety if you will, but banks, FI’s and fintechs have been paying a lot of attention to voice banking. Most of the bank-based virtual assistants such as BofA’s Erica, have voice banking capabilities. Yet, voice banking has a long way to go before it becomes totally efficient and secure. One major problem is that voice banking is still not sophisticated enough to understand the customer (whether it’s their speech or their personalized needs without having the user answer a plethora of questions). To achieve either of these requirements, voice banking lacks the intelligence to access and understand specific data. None of the voice chatbots currently out there have reached the sophistication required to make them reliable, frictionless banking channels.

And then, there is the issue of security. Recently, Bank Innovation reported on a kind of hack called dolphin attack, in which hackers can interact with systems by using inaudible sound frequencies, without the user knowing about it. So, if one of the most sophisticated voice bots aka Amazon’s Alexa is prone to such security problems, then how can banks be immune to it?