A new Paubox survey reveals that in healthcare organizations using AI coding assistants, IT staff build more often than developers, signaling a structural shift in who creates software. By Business Wire.

Paubox has released preliminary findings from a July 2026 survey of 151 healthcare leaders, revealing a significant shift in who builds software within the sector. The data, scoped exclusively to organizations already using AI coding assistants to handle protected health information (PHI), indicates that IT staff are the primary builders in 84% of surveyed organizations, compared to 58% for developers and engineers. Strikingly, in roughly two in five of these organizations, developers are not involved in AI-assisted building at all.

This trend suggests that the adoption of AI coding tools has widened the pool of creators beyond the traditional engineering function. The survey highlights that four out of five healthcare leaders cite the ability to let non-developers build software as a key driver for adopting these tools. Consequently, IT teams, operations staff, compliance officers, and even clinicians are increasingly producing software. This democratization of development is coupled with widespread deployment: 80% of the surveyed organizations have already implemented AI-assisted tools that patients use directly.

While the headline finding focuses on personnel rather than technology, the implications for governance are substantial. As non-technical staff generate more code, the burden of maintaining security and compliance falls on different organizational layers. Paubox notes that this is part of a broader benchmarking effort, with the full 2026 report set to detail the volume of tools produced and the agent infrastructure connecting AI services to core systems. Engineers should watch for how healthcare organizations manage the technical debt and security risks introduced by this decentralized development model, particularly as AI-assisted workflows become more integrated into patient-facing systems. Interesting read!

[Read More]

Tags ai-and-machine-learning business-and-emerging-tech software-engineering