Independent evaluation for AI products in K-12 education
AIEI is inviting a limited number of education technology companies to participate in independent, collaborative testing of AI products used with or for K-12 students.
AI products are already being used across classrooms and schools, but there are still few ways to test how they actually behave in the specific contexts in which students and educators use them. AIEI is developing an education-specific approach to product evaluation focused on identifying safety and ethics risks before they cause harm.
For this initial cohort, evaluation will be private and collaborative. Findings will be shared with the participating company and will not be published or shared externally. The goal is to give product teams rigorous, actionable evidence they can use to strengthen their systems.
What we can evaluate
We are currently accepting text-based AI products that can be systematically tested through an API or other instrumentable interface, including:
- AI tutors and student assistants
- Assessment and feedback tools
- Teacher copilots and instructional tools
- Student advising and support systems
- Safety and wellbeing tools
- Special education and accessibility tools
- Curriculum and content generation
Evaluations are designed around the product’s intended use, users, student age range, and deployment context.
What we test
Depending on the product and use case, testing may examine:
- Safety and harmful behavior
- Equity and bias
- Age and developmental appropriateness
- Accuracy and reliability
- Human oversight and escalation
- Student agency and over-reliance
- Privacy and sensitive disclosures
- Transparency
Testing combines systematic, automated evaluation with expert review where human judgment is needed.
How it works
Scope together
We work with your team to understand the product, intended use, users, technical architecture, and areas of greatest concern.
Test independently
AIEI conducts structured testing against an education-specific evaluation framework using repeatable test scenarios, adversarial testing, and other evaluation methods appropriate to the product.
Review together
We walk your team through the findings, including specific behaviors, areas of strength, and issues that may warrant attention.
Improve and re-test
Your team has an opportunity to make changes, and AIEI re-tests priority findings to understand whether those changes worked.
Independent evaluation. Collaborative remediation.
What participating companies receive
- A detailed evaluation of their product
- Review of findings with the AIEI evaluation team
- Identification of higher-risk behaviors and vulnerabilities
- Recommendations for addressing identified risks
- An opportunity to remediate findings
- Re-testing of priority issues following remediation
- A final evaluation report documenting results and improvements
Private and confidential
The evaluation is private and confidential.
Product-specific findings will not be published, ranked, or shared with schools, funders, customers, or other third parties without the company’s agreement.
Any technical access, documentation, data, or other non-public information provided for the evaluation will also be treated as confidential and used solely for the purpose of conducting the evaluation.
AIEI will not use participating companies’ proprietary technology, information, or access for any other commercial, research, or product-development purpose without their permission.
Participation
With philanthropic support, AIEI is able to cover a significant portion of the cost of evaluation for a limited number of products in this cohort.
Participating companies must provide the technical access and product information needed to conduct meaningful testing and commit time to review findings with the AIEI team.
Participation in the program does not constitute AIEI certification or endorsement. AIEI maintains independence over its evaluation methodology, testing, and findings.
Interested in participating?
We are currently identifying products for the next evaluation cohort. Tell us briefly about your product, who uses it, the age or grade levels it serves, how AI is used within the product, and why independent evaluation would be useful to your team.
About Just Horizons Alliance
The AI Ethics Index is an initiative of the Just Horizons Alliance, a 501(c)(3) public charity advancing responsible, human-centered innovation. Our work spans AI ethics, computational social science, simulation modeling, and the design of systems that strengthen human dignity, equity, and societal wellbeing.
Visit Just Horizons Alliance