AI Guidance for CPS
Students
In this section, we outline clear guidelines to ensure that our students uphold the values of CPS when using AI.
ITS
It’s essential to follow standard software engineering best practices to minimize potential bias and ensure data quality in GenAI models.
Privacy, Security, and Confidentiality
When using GenAI tools, it is crucial to understand that any information provided— whether through prompts entered by the user or AI-generated responses—could potentially be used by companies to train their models. This includes personal details of students, families, employees, and any proprietary or confidential information belonging to CPS. Assume that all information shared with a GenAI application will be used to train the model and could become accessible in the public domain.
To safeguard privacy and maintain confidentiality, you should:
- Never input personally identifiable information or protected health information into GenAI tools.
- Never input confidential, sensitive or legally protected information into GenAI tools. Legally protected information would include, though is not limited to, information contained in a student’s record; information that identifies students, employees, and other individuals; information contained in an employee’s record; proprietary information; etc.
- Never input copyrighted material or proprietary CPS intellectual property into GenAI tools.
- Staff: Contact your manager or the IT Service Desk at (773) 553-3925 if you are unsure whether information you are planning to input falls into any of the above categories.
Verification of the GenAI Tool’s Output
GenAI tools generate outputs based on their training from large and diverse data sets, which may include publicly available information. These outputs, while innovative, can contain fictitious elements (“hallucinations”) and may inadvertently include copyrighted or proprietary content.
To ensure responsible use:
- Always verify the accuracy and appropriateness of GenAI-generated content before sharing.
- Avoid using any GenAI outputs that might contain copyrighted material without clear ownership. This can be accomplished by, for example, examining the work for a copyright notice, considering the type of content and source (i.e. content issued by the US government is generally public), or referring to websites that store public domain works or the Copyright Database.
Bias and Fairness
AI models reflect the biases present in their training data, which can lead to outputs that unintentionally perpetuate stereotypes or discrimination. This is contrary to CPS’ commitment to diversity, equity, and inclusion.
To uphold our values:
- Conduct thorough reviews to ensure outputs are not only accurate, but also free of unintended biases and align with our educational goals.
- Verify and assess the source information that GenAI outputs are relying upon.
- Be vigilant of biases in GenAI outputs, particularly when these tools are used for decision-making or data analysis.
If you would like support in determining if GenAI outputs are demonstrating bias or how to correct for it, contact the Office of Equity at equity@cps.edu.