Dive Brief:
- Global banking leaders plan to use generative AI to expedite cloud migrations as board-level support builds, according to an NTT Data report published in February.
- Nearly two-thirds of banking institutions rely on mainframe infrastructure, the survey of 650 global banking leaders found. One in 5 haven’t started modernizing hardware.
- Despite legacy hurdles, the deployment race is on. More than 2 in 5 banks have integrated generative AI into tech stack already and around one-third have entered the early experimentation stages.
Dive Insight:
Banking CIOs have changed their view of the cloud.
Rather than seeking cost savings, executives are turning to cloud for flexibility, scalability and AI capabilities, according to the report. But technical debt is complicating adoption initiatives.
“Unfortunately, many of the banks are still working on code from the 1970s,” Michael Abbott, senior managing director and global banking lead at Accenture, said during a CIO Dive live event last month.
Generative AI capabilities are expected to ease digital transformations, even for large organizations weighed down by piles of tech debt. By 2027, Gartner predicts that generative AI tools will reduce modernization costs by 70%, according to analysis published last month.
Vendors have seized the opportunity. IBM trained watsonx to translate COBOL into Java and several other generative AI providers are focused on model coding enhancements.
Some banks are moving faster than others. Bank of America and Capital One own more than 70% of all AI patents filed by banking companies since 2010, according to industrywide analysis by Evident.
As banks progress toward their AI goals, CIOs will need to weigh the benefits as well as technology’s risks related to copyright, IP leakage and other challenges. CIOs who move too fast may just pile on additional tech debt, while those too slow to react relinquish their competitive edge.
Banks and other mainframe users aren’t aiming to move all critical workloads to the cloud. More than half of organizations prefer cloud for data analytics and reporting applications but want mainframes to run core systems and complex legacy applications, according to an Advanced report.
