How banks use artificial intelligence in risk management
2 min readArtificial intelligence can be used in bank risk management. The law obliges banks, for example, to establish an appropriate and effective risk management system that continuously ensures their risk-bearing capacity. The main aim is to identify market, credit, insolvency or fraud risks in connection with, for example, trading decisions or lending and to minimize these risks. Here in particular, AI or ML can help to recognize new patterns and thus contribute to minimizing risk. However, the financial regulator does not approve of individual algorithms. Rather, it examines the individual processes in a risk-oriented and ad hoc manner in their specific application in individual cases. However, when using AI overriding principles determined which must be taken into account by the financial institutions.
Management remains responsible for artificial intelligence and its use
Regardless of how sophisticated the artificial intelligence is, management remains ultimately responsible for the use of AI. This means, among other things, that management must have adequate technical understanding. If algorithm-based decisions are made, then risk management must also be adapted to these circumstances. This means, among other things, that the probability of damage occurring due to incorrect decisions made by the algorithm is analyzed and the results are documented. The same applies to the extent of the potential damage. Furthermore, an overarching framework is to be set up that specifically addresses the algorithm-based decision-making processes and takes into account their mutual dependency. If applications are obtained from external parties, management is also responsible for establishing effective outsourcing management.
No bias may be generated and legal requirements must be observed
When using AI, the systematic distortion of results (bias) must be avoided. Entrepreneurial decisions must not be based on bias. This should also eliminate the risk of damage to reputation if, for example, individual customer groups are disadvantaged due to the distortion. Companies are therefore required to use data of sufficient quality and quantity. In the development phase, financial institutions therefore have to develop a data strategy, for example, which ensures the permanent provision of data. The current data protection regulations must be observed. To ensure that the algorithms and models can be checked both internally and externally, there is a documentation obligation for the financial institutions.
