The main objective of this guide is to establish some criteria for the use of artificial intelligence tools: when and why we can use them, when we should not use them, some examples to illustrate how and for what we can use them, etc. The guide is just a starting point. Given that technology is constantly evolving, this is a document that we will update regularly.

We summarize below the main criteria for the use of AI tools that we have included in this guide:

• Always comply with the law and care for ethics. Current rules will always be respected. Under no circumstances shall personal and confidential data be uploaded to external systems that are not adequately protected or authorised.
• The work must always be supervised by professionals and the responsibility for this work is yours. The AI tools will be used as an aid, but the final decision and the responsibility of the work will always lie with the professional.
• Artificial intelligence tools should be used as an aid in the process, not as instruments that offer a final product. The responses of these tools will be considered contributions to the work, never final results. Therefore, Elhuyar will not publish any content generated by artificial intelligence (neither in its channels and means, nor in the content delivered to clients and collaborators).
• Ensure the correctness and appropriateness of the results in Basque. Many artificial intelligence models still have important limitations with the Basque Country. The texts will always be supervised by Elhuyar to ensure linguistic quality and thus remain a respected reference.
• Act transparently. When using an artificial intelligence instrument, this should be reported.
• Take into account the explainability. It is essential to know how the AI tools have obtained a given result and to know the origin and sources of information they have used to do so.

Full guide: Practical guide to the use of artificial intelligence tools