There is a growing call for the implementation of an AI-based screening system to identify unfair contract terms that shift costs and responsibilities onto subcontractors in construction contracts. With over 110,000 subcontracting agreements signed annually, it is challenging for human reviewers to assess each contract individually. The proposal aims to use AI to analyze contract documents and flag those with high risk.
According to the Korea Construction Policy Institute, approximately 116,000 subcontracting contracts are executed each year, based on national statistics. Each contract can consist of 30 to 50 pages, including cover pages, main texts, special provisions, and itemized statements, making it difficult to identify all unfair terms through current personnel-based reviews.
Unfair contract terms refer to conditions that allow the main contractor to exploit their superior bargaining position to limit the subcontractor's benefits or shift costs and responsibilities that should be borne by the main contractor. These terms can manifest in various forms, including project specifications, special conditions, agreements, and memorandums.
A survey conducted by the Korea Construction Policy Institute in 2022 revealed that 73.6% of specialized construction firms reported encountering unfair contract terms. The types of unfair terms included unrecorded cost shifts at 13.7%, additional work cost shifts at 12.7%, and delays in project completion cost shifts at 10.8%.
In decisions by the Fair Trade Commission, contract conditions that require subcontractors to set aside funds for defect repairs before project completion or to bear all compensation and expenses in the event of industrial accidents have been deemed unfair. Instances where main contractors withheld partial payments for completed subcontracting work were recognized as unfairly limiting subcontractor benefits.
Legal sanctions against unfair contract terms have been strengthened. The Subcontracting Act introduced a prohibition on unfair terms in 2013, and since April of last year, some unfair terms have been rendered legally ineffective. However, concerns about potential trade disruptions and the burden of proof have made it difficult for subcontractors to report these issues, and enforcement is often reactive.
A representative from the institute stated, "To address these issues, we need to establish a pre-screening system using AI. This would involve electronically collecting subcontracting contracts, special conditions, project specifications, and additional agreements, allowing AI to identify suspicious clauses related to cost shifts, responsibility shifts, and rights limitations, categorizing them by risk level."
The representative added, "If a contract is classified as high risk, the Fair Trade Commission can further review the contract details and the parties involved to determine whether an investigation is warranted. AI will not make the final decision on legality but will serve as a tool to narrow down investigation targets."
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.