Hurdling the barriers to AI implementation in every industry
While companies realise the importance of artificial intelligence in their area of work, there are still some misgivings about its purported ability to replace human beings.
From languishing outside the top 100 search terms on a leading IT research and advisory firm’s website in January 2016, artificial intelligence (AI) shot to the seventh place by May this year1. Clearly, AI is more than just the flavour of the season, with leading companies in every industry investing in its technologies – analytics, machine/deep learning, natural language processing, virtual assistants, chatbots, and robotic process automation.
AI fundamental to the success of a company’s strategy
Google, always a step ahead, announced its decision to go “AI first” sometime back2. That other enterprises will eventually follow is quite evident from the findings of an Infosys-commissioned study of 1,600 business and IT leaders, one of which was that 76 percent agreed that AI was fundamental to the success of their company’s strategy. The organisations that had already implemented AI, or were planning to, expected it would add nearly 40 percent to revenue by 2020. The survey also found that only nine percent of companies had no plans to implement AI.
How are banks and financial companies placed with respect to AI? Of the 10 vertical groups in the survey, financial services ranked eighth by maturity of AI adoption. Most banks were found to be “explorers” who were in the process of expanding their AI skills and planning to implement something in the next 12 months.
Understanding the barriers
These results are rather disappointing for an industry that is flush with data, and is powered by a huge array of IT systems and software. What are the barriers that are holding banks back from embracing AI despite a clear push from their top management?
The survey named “employee fear of change” – cited by 54 percent of respondents – the top barrier to AI adoption. This fear is even more acute in financial services where 25 percent of organisations felt that AI would result in redundancies, the highest among all verticals.
Shortage of implementation skills was in joint top spot. Incidentally, this finding is corroborated by a separate survey conducted by the IT research and advisory firm mentioned in the beginning of this piece. The only solution is that organisations contemplating AI deployment must acquire or nurture the necessary skills – development, workflow, production and systems skills – beforehand. A good way to do this is by collaborating with fintech firms and startups to tap into their talent resources.
The third barrier to AI adoption according to the Infosys survey, was a lack of clarity on where AI could be helpful, cited by 49 percent of participants. I believe banks do not have to look far to discover a variety of use cases. Pilot projects show beyond doubt that AI-led automation can yield huge benefits in terms of efficiency and cost reduction. This should be music to the ears of an industry that is under immense pressure to keep margins afloat. But the problem seems to be that financial service companies believe the cost of AI must come down before it can prove effective. They are also under the apprehension that the technology itself is not mature enough for adoption.
Early adopters stand to gain a significant competitive advantage
While it is true that AI is still developing (at a furious pace), it is sufficiently evolved to warrant deployment. The best way forward for banks is to set realistic expectations and move ahead. It is important not to hang back any longer because given the pace at which AI technologies are evolving, there is no chance for laggards to catch up later. Quick movers, on the other hand, stand to gain doubly – they get a head start at learning, and so do their AI systems. Some banks are building confidence by running pilot AI programmes within their organisation before taking the initiative to customers. A good example here is the Royal Bank of Scotland, which trialed a customer service bot for two months among staff, before allowing it to answer queries from customers.
Finally, there are barriers stemming from cultural issues, such as concerns about ceding control to AI, and facing resistance from employees. The important thing to remember is that AI is not here to make human beings redundant, but, rather, to enable them to do much more with their unique abilities. Banks should therefore see this as an opportunity to automate mundane, repetitive tasks so that they can devote their people to more valuable pursuits, such as creative thinking, problem solving, and innovation, which are currently outside the scope of AI. The good news is that the majority of organisations realize this: in the survey, 80 percent of companies planning to use AI were going to retrain the impacted workers and redeploy them in other areas.
This, really, is the best approach. Organisations don’t have much choice in the matter of adopting AI if they hope to survive into the future. Instead of viewing it as a fait accompli that is thrust upon them, they should look upon AI’s bright side, which is that it has enormous potential for differentiation, innovation, operational excellence and the amplification of human ability.
(Disclaimer: The views and opinions expressed in this article are those of the author and do not necessarily reflect the views of YourStory.)