On June 24, 2025, the District Court of New Taipei City rendered a judgment against the popular legal search engine Lawsnote, finding the legal-tech startup liable for copyright infringement by reproducing the description of legislative history created and compiled by Lawbank (111-Chi-Su-Zi No. 8).
Lawsnote was founded in 2016 by Mr. Guo, who is also an attorney, and aims to provide innovative search services for the legal industry. On the other hand, Lawbank, established in 1986, is one of the most prominent databank in Taiwan and known for its comprehensive legal research resources. According to the judgment, between 2018 and 2019, Mr. Hsieh, another co-founder of Lawsnote, developed and deployed six web crawlers to automatically extract data from the database of Lawbank, including the laws and regulations, the respective attachments and forms, and the summaries of legislative history. Lawbank discovered the unauthorized reproduction and filed lawsuit in 2021. The case primarily centers on two main issues: 1) whether the scraped content of legislative history is eligible for copyright protection; and, 2) whether Lawsnote’s use falls under fair use. The Court ruled against Lawsnote on both.
On the issue of copyrightability, Lawsnote argued that the legislative history it downloaded, which totaled 98,068 files, should not be eligible for copyright, and that even such description is copyrightable, based on doctrine of merger, Lawsnote should not be liable. The Court disagreed, elaborating that the threshold for copyright protection is low, and that Lawbank’s selection, arrangement, and presentation of the legislative history demonstrated minimum degree of creativity that could meet the requirement. As to the defense of merger, the Court rejected Lawsnote’s arguments, noting that other legal databases described the legislative history differently, showing expressive choices were not limited.
On fair use, the Court weighed the following factors: the purpose of the conduct; the nature of the work; the amount and substantiality of the work used; and the impact on the market. Citing the fair use analysis in Thomson Reuters Enter. Ctr. GmbH v. Ross Intel. Inc., the Court found Lawsnote’s use was commercial in nature. By using crawler to download the bulk data from Lawbank, Lawsnote copied and reproduced 100% of Lawbank’s copyrighted content.
Lawsnote argued that the policy should favor public access to the content of the law, and that its search-focused platform was different from the model of Lawbank. The Court dismissed these arguments, reasoning that Lawsnote gained an unfair competitive edge by avoiding years of efforts and costs in compiling the data, which harmed Lawbank’s market. Thus, the Court found Lawsnote’s conduct did not constitute fair use, and Mr. Guo, Mr. Hsieh, and Lawsnote should be jointly liable for infringing Lawbank’s copyright.
While the judgment is not final and it remains uncertain how the case will evolve, this case nonetheless offers several noteworthy insights regarding copyright infringement issues.
First, one should always verify the source of the content. Using licensed materials may be time-consuming, but safer. Second, fair use is less likely to apply where there is mere wholesale copying without any creative contribution. Lastly, while this case does not involve using copyrighted materials for AI training, it provides a useful preview of how courts may approach future disputes over bulk data scraping. The purpose and character of use, and the impact on the original market will likely remain the most critical fair use factors.