Banks Without Strategy in Artificial Intelligence
According to a recent study, the financial sector is intensely concerned with technologies from the field of artificial intelligence (AI). However, most institutes are still in a preparatory or test phase and cannot demonstrate a coherent strategy.
According to the results of the study, almost all financial institutions have already dealt with the topic and have high expectations of it.
AI is complex, however, because there are different technologies behind it. Only a few institutes already use AI in regular operation, most of them are still in the evaluation phase.
AI can reduce error rates and significantly reduce response times to customer requests. In addition, employees can be freed from tedious routine work and given freedom for value-adding activities.
The expectations on the part of the banks are correspondingly high. The respondents expect AI to have advantages in competition and in process optimization, followed by economic improvements, ie increasing profits and falling costs.
The greatest potential for AI is seen in the back office (60 percent), followed by products (50 percent) and customers (47 percent). At 38 percent, chatbots are the most common AI solution used by financial service providers, followed by machine learning (32 percent).
Behavioral use of AI
Three-quarters of those surveyed see themselves as part of an early majority in artificial intelligence, almost 30 percent of whom describe themselves as early adopters or even innovators. Only a quarter belong to the late majority or the laggards.
However, only twelve percent already have up to three AI applications in use; ten percent have more than three. Twelve percent of the institutes surveyed are in the phase of building up know-how, 22 percent are considering specific uses of AI technologies, 18 percent have developed use cases and 24 percent are currently testing prototypes.
The majority of banks and capital management companies surveyed stated that they had already introduced at least one AI solution, followed by commercial banks and savings banks. Cooperative and state banks are in the middle alongside fintechs, service providers and insurers. Promotional and private banks are at the bottom.
Missing strategy for artificial intelligence
Not one of the institutes involved has a defined and fully implemented AI strategy. Even institutes that have already used AI occasionally did not pursue a consistent strategy, nor would they have managed to anchor robust and stable processes, procedures and methods in the organization.
32 percent state that AI is not part of the corporate strategy. 53 percent state that the strategy is currently being worked out or that implementation is planned. A minority of 15 percent say that the AI strategy is at least partially implemented.
There is also an urgent need to catch up with the introduction of AI governance — the anchoring of corresponding processes, methods and procedures in the organization. 62 percent of the respondents state that they are not being introduced in an appropriate form, 28 percent have defined the first standards and only ten percent are already working on them or are in the process of optimizing the processes.
In addition, the institutes will soon reach their limits when introducing AI. According to their own assessment, the availability of resources is the greatest challenge and is mentioned by 54 percent as a particularly pressing bottleneck. This is followed by a lack of know-how and poor data quality (42 percent each).
As a result, investments in internal and external resources are urgently needed. The companies surveyed are also aware of this and plan a median of two to five percent of their IT and project budget for AI investments for 2020.
At the same time, there are high demands on the efficiency of AI. AI has to achieve six to twelve percent profit or cost-saving potential for the majority of the respondents so that a positive implementation decision can be made. However, realizing this is likely to be difficult — also in view of the current level of maturity with regard to AI.
Author: Marko Vidrih
Featured image credit: Pixabay