Highway QA/QC relies on records such as test results, inspections, certifications, and photographic evidence to determine whether materials and workmanship meet contract requirements. These records support material acceptance, contractor payment, and infrastructure decisions, so integrity matters. Recent federal reviews found weaknesses in how highway QA/QC programs are guided and reviewed. Separate cases show that test results can be falsified or manipulated without AI. Following AI adoption in highway projects, more studies have focused on using AI for highway inspection and defect detection than on the risks it introduces. AI can generate, edit, or process QA/QC evidence, including altered test files, edited technical images, and documentation that lets users bypass specifications and review procedures. This research examines AI-induced risks to QA/QC data integrity in highway construction using a Bibliometric-Systematic Literature Review (B-SLR), which combines quantitative mapping of the field and a close reading of its content. Documents were collected from multiple databases and screened through a defined process. The bibliometric mapping shows research on AI in highway construction is recent and thematically fragmented, with little consolidation around quality and integrity concerns. The systematic review builds on this, tracing how AI has moved into highway construction and examining the risks it introduces to QA/QC evidence, to identify current research gaps and federal needs. Next steps include a structured survey of state DOTs and FHWA, followed by expert validation. The end goal is to build intelligent software that detects, assesses, and mitigates AI-induced risks in QA/QC to support trustworthy highway construction.