Tool Approval and Integration
- All AI EdTech tools provided by vendors must undergo the same rigorous vetting and approval process as all EdTech tools known as the EdTech Request for Qualifications (RFQ) by CPS before they can be used by students or with student data. This process is currently available once a year and vendors can sign up for notification of future contract opportunities on the CPS Procurement Website.
- Approval signifies that the tool meets the District’s standards for privacy, security, and confidentiality in compliance with all applicable federal, state, and local laws (FERPA, COPPA, SOPPA, etc.) and District policy.
- Vendors must provide comprehensive documentation regarding their tool's data privacy practices, including how data is collected, stored, used, and protected. This information will be a key factor in the approval process.
- Vendors should be prepared to provide information regarding the age appropriateness of their tools, including any age restrictions outlined in their Privacy Policy or Terms and Conditions.
Privacy, Security, and Confidentiality
- Vendors must ensure that their AI tools and services are designed and operated to protect the privacy and security of all CPS stakeholders, including students, families, and employees.
- It is imperative that no personally identifiable information (PII) or protected health information is collected, stored, or processed in a manner that goes against CPS guidelines and could compromise the privacy of our stakeholders.
- Vendors must ensure that stakeholder information, copyrighted material, or proprietary CPS intellectual property is not used to train their models without explicit permission.
- Vendors should implement robust security measures to prevent unauthorized access, data breaches, and misuse of CPS data. Security by design principles should be followed, with continuous evaluation through testing, patching, and authentication measures.
Ethical and Responsible Use
- Vendors are expected to develop and deploy AI technologies with a steadfast commitment to fairness, transparency, and accountability.
- Efforts should be made to minimize potential bias in AI models, recognizing that biases in training data can inadvertently perpetuate discrimination, which is contrary to CPS’ commitment to diversity, equity, and inclusion. Vendors should strive for algorithmic fairness and be transparent about the potential for algorithmic bias.
- It is critical to address the issue of hallucinations, where AI models produce false or misleading information. Vendors should have mechanisms in place to mitigate these inaccuracies and clearly communicate this potential risk to users.
- Vendors should strive for explainable AI, providing insights into how their systems work and how decisions are made while promoting a human-in-the-loop approach, recognizing that human oversight is essential for responsible and effective AI use.
Support and Collaboration
- Vendors are expected to provide adequate support and resources to CPS staff for the effective and responsible use of their AI tools. This includes providing training materials, technical support, and ongoing updates.
- Vendors should maintain open communication channels with CPS to address any concerns, issues, or updates related to their AI offerings. This includes collaboration with CPS on professional development initiatives related to their specific tools.