Do LLMs Favor Their Providers? Measuring Vertical Integration Bias in Code Generation
This work identifies and measures a new form of bias in LLM code generation that could constrain developer choice and increase provider lock-in, relevant to the AI and software engineering communities.
LLMs affiliated with providers exhibit a bias toward their own ecosystem in code generation, with effects up to +18.8 percentage points in direct generation and +39.2 pp in agentic workflows, and early biased choices can persist in downstream files with up to 90.3% persistence.
Large Language Models (LLMs) have become an integral part of software development, especially with the advent of agentic capabilities. Yet, many frontier LLMs are affiliated with specific providers. This raises the question of whether generated code favors the provider's own ecosystem over comparable alternatives, potentially constraining developers' choices and increasing dependence on a single provider. We define this behavior as Vertical Integration Bias (VIB) and introduce \textsc{VIBench}, a benchmark for measuring VIB in direct and agentic code generation across $20$ provider-selectable software-integration scenarios. Evaluating $10$ frontier provider-affiliated models against $3$ non-affiliated controls, we find positive VIB in direct generation, with six of ten affiliated models showing statistically significant effects up to $+18.8$ percentage points (pp). Agentic workflows further amplify VIB, reaching $+39.2$ pp. Moreover, early affiliated-ecosystem choices in agentic workflows can persist into conceptually decoupled downstream files, with persistence as high as $90.3\%$. These findings underscore the need to measure and account for VIB in code generation, especially as agentic capabilities become more prevalent.