Centre interdisciplinaire sur les enjeux stratégiques (CIENS)

AI Charter

This charter aims to regulate the use of AI in an ethical, responsible and sustainable way for CIENS students but also its teachers and researchers. It explains its operation, warns of its harmful effects for the environment, defines the conditions of authorized use, prohibits and exempts recommendations on transparency.  

How does IA work?

Text generation tools, based on language models (LLM), are trained to predict the most likely word to form a consistent and relevant sequence. Algorithms are used to learn, in particular from existing data on the web, to generate new content (texts, images, music, codes, etc.).

an IAg is far from being 100% reliable and has some malfunctions, including:

  • Hallucinations: due to their internal functioning (based on statistics), IA tools produce plausible content, but which can be partially or entirely erroneous. The term “hallucination” is used to designate generated content that is false, which requires a systematic verification of content on the part of the user.
  • Biases: the content generated by IA can reproduce the representational biases present in the data used to form these tools. Users must be aware of the risks of reproduction and amplification of pre-existing biases and exercise precautions.

Environmental impact 

lmpact is at 4 levels:

1) Data centers house electronic materials and therefore raw materials from rare earths, often extracted in a destructive way for the Environment.

2) Data centers produce large amounts of electronic waste (containing hazardous substances, such as mercury and lead).

3) The construction of these data centers and the cooling of electrical components require huge amounts of water. As a reminder, a quarter of humanity has no access to drinking water.

4) The technology of AI requires a lot of energy from fossil fuels producing greenhouse gases responsible for global warming.

Authorized uses

These tools can be an ad hoc help in certain stages of the work such as translation, documentary exploration, exploratory analysis of documents or work planning. However, they cannot replace the intellectual effort and the personal reflection that must be provided.

Prohibited Uses

  • Production of false references
  • Develop a scientific problem and reasoning
  • Write text content
  • Provide to theIAG Sensitive, protected and confidential information.

In general, always ask yourself if using AI will add value to your work or if it is an easy way to avoid thinking and therefore training yourself intellectually (which is the reason why you registered for our courses).

Transparency Requirement 

Any use of an artificial intelligence tool must be mentioned by a note or in the body of the text; The tool used as well as the nature of the use made of it must be specified. Users are responsible for systematic verification of the accuracy of the information generated by AI. No information provided by AI can be considered as a scientific source.

The undeclared use of IA is likely to be considered an academic fraud. Disciplinary sanctions may be applied.

Sources used for this Charter: