This study aims to explore which factors influence public procurement professionals’ acceptance of Generative artificial intelligence (GenAI) in operational procurement tasks. While GenAI has potential to automate routine work, little is known about how procurement professionals experience its practical value, and under which conditions they accept it.
The authors conducted 31 semi-structured interviews with municipal buyers and procurement consultants across two rounds (2024 and 2026). This two-round design captures shifts in perceptions as GenAI matured and organizational policies developed and confirms findings across a broader sample. Thematic analysis identified four key themes explaining GenAI acceptance.
Professionals view GenAI as helpful for administrative tasks, such as drafting tender documents and contracts. However, trust and human oversight are seen as essential: professionals want to retain responsibility for final decisions. Some worry that overreliance on GenAI may erode their critical thinking, particularly among newcomers. Professionals are experimenting on their own, but unclear policies, limited management support and regulatory uncertainty hold back broader acceptance. The second round revealed that while organizations provide approved tools, a gap persists between formal policy and actual practice.
The study is based on Dutch municipal procurement professionals and may not generalize to other countries or private sector contexts. Future research could test the identified factors quantitatively.
Acceptance goes beyond individual willingness. Procurement-specific training and clear guidelines can support responsible use and reduce administrative burdens without undermining professional judgment or public accountability.
Overreliance on GenAI may weaken the critical judgment needed to protect public interests, particularly among younger professionals. Clear policies and targeted training are essential to ensure GenAI supports public accountability.
This two-round design shows that GenAI acceptance depends on trust, human control and organizational conditions (not just task fit and usefulness), offering new perspectives on GenAI implementation in the public sector.
