The EU AI Act imposes extensive compliance obligations on organisations that develop, deploy, or integrate AI systems, covering security, privacy, data governance, transparency, accuracy and robustness. Small‑to‑medium enterprises and individual developers often lack the legal expertise to interpret these obligations and translate them into engineering and governance practices, creating implementation gaps and the risk of missing safeguards or misdirected development. To bridge this gap, a variety of automated EU AI Act compliance checkers (AIACCs) have appeared, claiming to streamline assessments and provide practical guidance. This paper presents the first empirical study and assessment of AIACCs, analysing twelve mainstream tools across dimensions such as interaction mode, user‑friendliness, legal coverage, report structure, determinacy and actionability. The findings show that:
- Quality varies widely; current tools are best suited as early‑stage orientation aids;
- Interaction patterns and usability are inconsistent, with many tools oversimplifying or omitting key obligations;
- Generated compliance reports often lack deterministic, actionable guidance, fostering a false sense of compliance.\ Based on these observations, we provide a critical baseline of the AIACC landscape and propose design principles for more reliable compliance‑support tools.\
Review: This study delivers the first empirical evidence on regulatory‑tech tools for the EU AI Act, warning practitioners to use automated checkers cautiously while offering clear directions for future tool improvement.