<![CDATA[Background. The virtualization of sign language through artificial intelligence in social media platforms presents linguistic challenges that have not been widely explored, especially related to the accuracy of syntactic structures in digital contexts. These visual representations have the potential to reproduce grammatical misconceptions that impact the meaning and effectiveness of communication. Purpose. This study aims to explore how the syntactic structure of sign language is represented in a virtual format by AI systems used in social media such as TikTok, Instagram, and YouTube, as well as identify their accuracy and distortions. Method. The research uses an exploratory qualitative approach with a cross-platform case study design. Data were obtained from 30 virtual sign language videos and analyzed using visual-spatial linguistic frameworks and open coding techniques. Validation is carried out through thematic triangulation analysis and expert consultation. Results. The results show that the representation of syntactic structure varies greatly between platforms, with YouTube being superior in accuracy to TikTok. Factors such as the length of the video, the sophistication of the algorithm, and the presence of non-manual elements greatly affect the completeness of sentence structure in virtual sign language. Conclusion. The current representation of sign language by AI does not fully reflect the complex syntactic structure. A new approach is needed in the development of multimodal-based technologies that consider linguistic elements as a whole to make digital communication more inclusive and accurate.]]>
Copyrights © 2025
