AI could eliminate up to 70% of manual banks' risk tasks

Artificial intelligence could eliminate 60-70% of manual operational tasks in banks' risk functions, allowing lenders to redirect resources towards deeper risk analysis, strategic decision-making and AI development, according to a report by McKinsey.

The report said banks could use AI to automate routine activities such as data gathering, documentation, compliance checks and monitoring, while retaining human oversight for complex assessments and important decisions.
McKinsey said banks could reinvest half of the capacity released through automation into stronger analytical capabilities, strategic risk advisory and faster portfolio decisions. The remaining capacity could be directed towards specialised AI engineering, data governance and platform development.

The report said, "We estimate that AI-enabled workflows could eliminate 60 to 70% of manual operational tasks, depending on the process and degree of automation."
It added that selected risk workflows could achieve structural efficiency gains of 30-60% after accounting for human validation and oversight.
Some leading banks are targeting productivity improvements of significantly more than 30% across their risk functions over the next two to three years, compared with the 5-10% annual targets common under traditional approaches.

AI is already being used in areas such as credit assessment, financial crime monitoring, regulatory compliance and risk management. In wholesale lending, early implementations have delivered productivity improvements of 30-50% across the credit process, with output quality rates of around 95%, the report said.
The technology can also help banks identify potential borrower stress earlier by continuously monitoring financial information, credit ratings and sector developments, allowing lenders to reassess exposure before problems become more serious.

However, McKinsey cautioned that scaling AI would require changes in workforce planning, technology systems and governance. Banks would need skilled employees who understand both risk management and AI, along with clear rules to check the accuracy and reliability of AI-generated outputs.
The report said fragmented AI projects could fail to deliver wider benefits unless banks build shared technology platforms and common governance standards.

"The transition to an AI-enabled risk operating model is not a speculative question of if, but an operational mandate of when and by whom," McKinsey said, adding that early adopters could gain cost and productivity advantages through faster and more informed decisions. - ANI