AI coding tools have made it remarkably easy to build software modeled on commercial products. With “vibe coding,” a term coined by AI researcher Andrej Karpathy in early 2025, a user describes what they want in plain language. Then an AI assistant writes, tests, and revises the code, often with little review of what’s underneath.
That means a business analyst with no engineering background can now produce a working version of a CRM dashboard, a project tracker, or a field-service app in a single afternoon. That has ISVs worried about being copied — and not just by competitors but also by their own customers, who may decide to build a “good enough” version in-house rather than renew a subscription.
Although enterprises will likely keep demanding compliance, security testing, and ongoing support, which home-grown apps rarely provide, “good enough” at a lower cost is a powerful motivator, especially for smaller organizations and departmental tools. As a result, ISVs may shift value toward what’s difficult to replicate: data, ecosystem integrations, partner services, and accountability when something breaks.
It’s a trend worth watching as AI reshapes how products get made. Channelnomics CEO Larry Walsh shares his thoughts from London in the latest episode of “In the Margins.”