artificial intelligence


By Dan McKinney, Co-Founder and Chief Marketing Officer of Finxact

Artificial intelligence is a buzzword gone bad. The multitude of breathless science fiction-like stories about killer robots, a collapse of the job market and highways full of self-driving cars has obscured the simple utility that specialized AI can provide today.

The core concepts of AI have been around for decades, growing at academic institutions like Carnegie Mellon University, MIT and Stanford, then maturing rapidly when paired with the missing component: massive amounts of data. By pumping petabytes of data into their machine learning algorithms, data-driven companies like Facebook, Google and Amazon have accelerated the pace of advancement and made AI real to the common man.

For banks, AI is not a singular tool, but rather a “Swiss Army knife” of applications that can help banks improve processes and add services. It’s not a massive behavioral shift similar to what internet and mobile banking did to the customer relationship. Instead, it’s a catalyst for banks to be more precise across all aspects of their business. AI can provide better underwriting, stronger cybersecurity and fraud protection, more tailored customer interactions, and faster, more efficient operations. By operating at a higher level, banks can eliminate costs and raise revenues in ways that customers may never perceive.

As banks start to bring AI into their business, it’s important to consider the Swiss Army knife analogy again. The base of the knife is not some omnipotent AI system, but instead the core banking platform where the data, applications and algorithms come together in a series of APIs and integrations. AI systems for the foreseeable future will be specialized for certain tasks. Just as you wouldn’t expect anti-virus software to become a spreadsheet, a virtual chatbot won’t be performing compliance checks. Instead, the next level of AI’s impact in banking will be reached by AI systems interacting with other AI systems. This puts the burden on a bank’s central systems to be open and flexible to allow for speedy and seamless data access for old and new systems alike.

Ultimately, AI-powered precision will generate benefits that reinforce themselves across the different processes of a bank. For example, if a bank has higher confidence in its loan portfolio because of underwriting techniques that reduce the risk of default by a few percentage points, it could lower its interest rates to attract new loans, knowing that the growth in accounts with lower risk will outpace the smaller margin per loan. That’s single system thinking though. What if instead, a bank used its safer loan portfolio to be more opportunistic in the use of its liquidity to drive gains.

AI is capable of elevating banks that fully invest in the components to make AI work best:

  • Internal and external data sources
  • Agile machine learning algorithms
  • Real-time systems (up-to-the-second data and rapid reactions)
  • Innovative minds, from data scientists and business analysts to application architects

By building a platform for AI systems, banks can perfect their operations in the next decade and be a central pillar in society in ways that fintechs and non-banks can’t match.