Techno-linguistic bias in machine translation can occur when a machine translation system is trained on one language or text type, leading to inaccurate translations in other languages or text types. This can happen due to variations in training data representation or the presence of language-specific terminology and concepts. The implications of techno-linguistic bias are significant, especially for low-resourced languages, endangered languages, and dialects. Machine translation systems must take into account the characteristics of different languages and dialects during training to avoid inaccurate translations that could harm the preservation and transmission of cultural heritage and linguistic traditions. It is crucial to develop machine translation systems that cater to the diversity of languages and dialects, providing accurate and respectful translations for all users. We propose a method for assessment of techno-linguistic bias in MT and GPT-based services using a multilingual parallel corpus.