GenAI's Double-Edged Sword: Accelerating Maintenance, But Organizational Hurdles Remain

Steve Smith, Global Head of Modernization and Platforms at Equal Experts, has highlighted the profound impact of Generative AI (GenAI) tools such as Cursor, Claude Code, and Copilot on software maintenance. Smith, recalling early demonstrations by co-founder Ryan Sakorski, witnessed these tools generating application code, configuration, deployment pipelines, and telemetry within minutes. This capability signals a significant shift in how non-differentiating work, including bug fixes, platform upgrades, and security vulnerability patching, can be handled, potentially liberating developers from what Smith metaphorically terms ‘salt mine’ tasks and fundamentally altering the economics of software development.

However, the widespread success of GenAI in accelerating maintenance is not solely a technical achievement but an organizational challenge. Equal Experts’ Chris Rutter’s research indicates that while GenAI can dramatically boost coding speed, its overall impact is constrained by slower downstream assurance activities like end-to-end testing and change approvals. Enterprise organizations face hurdles in tool procurement, establishing robust security guardrails, and navigating misaligned incentives across various teams in the delivery pipeline. Smith emphasizes that true end-to-end acceleration requires a strategic reevaluation of these often-manual or semi-automated processes. GenAI can facilitate the automation of test creation, simplify domain understanding, and enable the adoption of ‘standard changes’ for routine deployments, thereby transforming what once took weeks into a matter of days. Ultimately, the ability of GenAI to reduce remediation time for maintenance tasks depends on an organization’s willingness to address these foundational process and cultural challenges.