Programmers are embracing AI coding tools, though some remain skeptical of their accuracy. FinServs should deal with AI-powered programming proactively.

Recent surveys from GitHub and Stack Overflow show that AI-powered coding tools are making a big impact.  

GitHub partnered with Wakefield Research to survey 500 US-based enterprise developers and found that 92% are using AI coding tools and that 70% believe that AI is providing significant benefits to their code. More specifically, AI tools are helping them achieve improved code quality, faster outputs, and fewer production-level incidents. 

A Stack Overflow survey of ~89,000 global developers, meanwhile, showed similar uptake with more nuance around trust. The survey found that 70% of programmers were already using or planning to use AI coding tools this year, but only around 42% said they highly trust or somewhat trust those tools.  

The results of both surveys highlight why FinServs should be taking a proactive approach in guiding experiments with their developers: Because if they don’t, the engineers will probably use AI tools anyway.  

By setting up pilot projects for tracking the accuracy, safety, and efficiency of tools like GitHub Copilot, Tabnine, and AWS CodeWhisperer, FinServs can protect themselves and get a concrete understanding of benefits.  

For example, Goldman Sachs started proactively setting up experiments that allow engineers to use generative AI tools to write and test code more efficiently and is tracking the progress.