Managing AI Capability Differences and Pay Transparency Requirements

AI-aided productivity is beginning to raise a question most compensation structures were not built to answer: should measurable differences in AI-enabled work affect how employees are paid? As AI use becomes routine across professional roles, organizations will need practical methods to determine when sustained differences in work should influence pay — particularly in states that have adopted pay transparency requirements.

Over the past year, our consulting practice has routinely asked clients whether an employee’s use of generative AI has ever resulted in higher pay. Almost without exception, the answer has been no. This does not suggest that employers have ignored AI. On the contrary, most organizations reported having policies governing the use of AI tools such as ChatGPT, Gemini, and similar applications. Likewise, organizations whose primary business involves developing AI technologies, or whose workforce includes technical AI professionals, generally have compensation programs that recognize those specialized skills. The distinction appears elsewhere. Among the vast majority of professional positions — accountants, budget analysts, human resource professionals, project managers, marketing specialists, and countless other non-technical roles — formal compensation practices that recognize AI-enabled work remain uncommon.

The absence of such policies is understandable. Until recently, differences in work produced using generative AI were neither widespread nor significant enough to warrant consideration for compensation. Most employees either did not use AI or used it only occasionally, so the differences in their work were relatively small. However, as the practical use of generative AI continues to expand, that assumption is becoming increasingly difficult to sustain. While organizations have developed policies governing whether employees may use AI, far fewer have considered whether sustained differences in AI-enabled work should eventually influence compensation. That distinction, while subtle today, is likely to become increasingly important over the next several years.

The issue can be reduced to a relatively simple question.

If the answer is no, then an equally important question follows. What objective process should employers use to distinguish those differences, and by what measure should they determine whether the resulting compensation differential is justified?

The significance of that question becomes more apparent when viewed within the broader context of pay transparency. Although the specific requirements vary by jurisdiction, pay transparency generally requires employers to disclose compensation ranges while equal pay provisions require that objective, work-related factors support those ranges. Historically, employers have relied upon factors such as experience, education, training, job complexity, responsibility, and sustained performance to explain why one employee may be compensated differently from another. These principles are well established and, in most organizations, provide the foundation for compensation structures.

The issue is not whether employees use AI. That question has largely been answered. Rather, the issue is whether sustained differences in AI-enabled work ultimately become comparable to other work-related factors that employers have traditionally considered when setting compensation. If AI consistently changes the quality, scope, complexity, efficiency, or business impact of work, then compensation systems will eventually need to determine whether those differences should remain solely within the performance management process or whether they have become sufficiently material to warrant consideration as part of the compensation structure itself.

That question is the starting point for the practical framework this series will build.