Generative artificial intelligence at HEC Montréal

Preamble

The School takes a balanced, coordinated approach to the use of generative artificial intelligence (GenAI). It encourages innovation by considering the potential risks and benefits. In keeping with the School’s mission, vision, and values, any use of GenAI tools must be done in compliance with the legal and internal frameworks in force.

HEC Montréal has signed the Montréal Declaration for Responsible Development of Artificial Intelligence. It adheres to the ten principles and ethical values set out in the Declaration.

1. Subject 

1.1 These guidelines are aimed at HEC Montréal students and employees and are intended to promote the responsible use of GenAI tools. 
 
1.2 They are not intended to prohibit or limit the use of GenAI tools, but rather to identify what needs to be taken into consideration in order to develop good habits and practices, depending on the context in which they are used. 
 
1.3 They cover the use of GenAI tools that are included in the Microsoft environment, as well as some that are readily available and offered by entities outside the School, which the School has not purchased. 

2. Security and protection of personal information and institutional data

2.1 Many GenAI tools collect, use, communicate, and store user data. It is important to have a clear understanding of what data is affected for each specific tool, to reduce the risk of sensitive or personal information being used in an unintended or unauthorized way. 
 
2.2 Employees must not, without prior authorization from the Information Access and Privacy Committee, enter into a GenAI tool any personal information that has been entrusted to HEC Montréal for institutional purposes. Entering personal information into GenAI systems without prior authorization may be a violation of privacy and of HEC Montréal’s operating rules.
 
2.3 Here are a few information security recommendations when using GenAI tools: 

• Do not enter any confidential, personal, or sensitive information into a GenAI tool, unless a thorough data security, risk, and privacy assessment has been conducted on that tool. For more information on the evaluation process, see section 9.  
• Opt for GenAI tools that have been evaluated, verified, and approved by the School and follow the terms of use.
• Before creating an individual account with a GenAI tool provider, take the time to read the terms of use and confidentiality policy.  
• Be on the lookout for counterfeit products, and make sure you only access the GenAI tool via its official website.
• Make sure that network communication with the GenAI tool is encrypted, and always check the URL in the address bar.
• Keep an eye out for updates, improvements, and possible limitations to the GenAI tool, and consult the latest user guidelines.  
• Be aware that GenAI can be used to create sophisticated phishing emails and "deepfakes," which are videos or audio recordings that realistically imitate a person’s voice or appearance without their consent.
• Be vigilant and report any information that could suggest an information leak or the publication of misleading, false, or defamatory information. For more information on the resources to contact, see section 9. 

3. Copyright and intellectual property

3.1. When entering information, it is crucial to exercise caution if it includes information protected by intellectual property rights. Users should obtain the necessary authorizations before using such information.

3.2. People using GenAI tools need to be aware of the ongoing concerns about copyright and ownership of the data used to train language models. Depending how these issues evolve, the terms of use for AI-generated content could change.

3.3. As the law currently stands, there is no consensus on copyright ownership of a work generated using GenAI.

4. Transparency and responsibility

4.1 Anyone using GenAI tools must do so transparently and in compliance with the applicable rules for mentioning and citing, so that everyone is informed of how they were used to generate content or carry out tasks.  
 
4.2 When using GenAI tools, we recommend taking a diligent and proactive approach by considering the limitations of these tools and our responsibilities when we choose to use them. 

5. Reliability and validity

5.1 GenAI tools may provide false, inaccurate, or incomplete results or content due to the inherent limitations of this technology. It is, therefore, crucial to check GenAI results and content against reliable sources before using or sharing them. If it is not possible to verify the results or content, you should not use them.
 
5.2 GenAI tools can also produce computer code that is not in compliance with the secure development framework and that may contain vulnerabilities or malicious code.

5.3 These vulnerabilities can lead to unauthorized data access, system intrusions, sensitive information leaks, etc. It is crucial to verify and approve generated content before using it. 

6. Bias management

6.1 GenAI models can reflect implicit prejudice in the data on which they are trained, which can lead to biased or discriminatory results.

6.2 It is highly important to promote ethical and responsible use of GenAI tools by adopting measures to identify, mitigate, and correct bias in generated content.  

7. Access equity

7.1 It is important to be aware that some people at the School may have unequal access to these technologies.

7.2 Some members of the community may not have the same resources or skills to use these tools optimally. It is important to adopt an approach that is sensitive to existing disparities and, if needed, to take them into account in our interactions.  

8. Exploration

8.1 Developing and using GenAI must be done by exploring the related issues and possibilities. Collaboration and sharing are essential to deploying, absorbing, and creating collective knowledge. For this reason, HEC Montréal provides its entire community with free access to Microsoft Copilot. 

9. Resources and support

9.1 Technical support or acquisition of an AI tool 
To receive technical support or purchase an AI tool, contact the IT Service Centre (in French). Please note that support is prioritized for HEC Montréal employees.  
 
9.2 Security evaluation 
The IT security team provides assistance and advice for using GenAI tools. It evaluates data security, identifies potential risks, and checks and approves the use of GenAI tools. Contact the IT Service Centre (in French).
 
9.3 Information leak or publication of misleading, false, or defamatory information 
To report such an incident, contact the IT Service Centre (in French).  
 
9.4 Using GenAI tools for teaching or research 
For any questions, contact the Learning and Teaching Innovation Centre: daip@hec.ca. 
 
9.5 Identifying, selecting, and organizing information 
For help identifying, selecting, and organizing information that can be used with GenAI tools, contact the Institutional Information and Archives Management Service: archives.info@hec.ca.