Knowledgeworker Create supports the automatic translation of content, leveraging artificial intelligence models. This article provides an overview of the services used and the potential risks of applying artificial intelligence in translation. Discover how to make optimal use of these technologies while ensuring responsible application.
Risks and responsible use of AI
Although artificial intelligence technologies offer many advantages, there are potential risks:
- AI models may incorrectly translate terms if they were not part of the training dataset. This can lead to inaccuracies in specific technical jargon. Therefore, content should either be reviewed or users should be made aware of the risks.
- Content sent to external services for translation might be sensitive. The content to be translated should not include personal data.
- AI models may introduce unintended biases into translations, which can be problematic for localizing content. Here as well, reviewing the translated content is recommended.
Services used in Knowledgeworker Create
Depending on the use case and language, Knowledgeworker Create utilizes various services to translate content. The combination of multiple services enables powerful translation of diverse content types.
The following services are used for KAI translations:
- DeepL API Pro (translation of knowledge modules and other content types, document translation)
- OpenAI (translation of knowledge modules and other content types, translation of rarer languages, translation of complex content types and formats)
Additionally, you can integrate custom translation services into Knowledgeworker Create, e.g. to leverage existing contracts or integrate your organization’s translation memory systems.