Think Tech legal and district documentation set
This policy describes how Think Tech approaches the responsible development, deployment, and use of artificial intelligence within educational environments. It applies to any Think Tech feature that uses AI, machine learning, large language models, or similar automated technologies.
1. Purpose and scope
AI Features are intended to support Educators and Educational Institutions by assisting with instructional preparation, lesson creation, task or rubric generation, summaries, recommendations, analysis, and other school-authorized educational or administrative functionality. AI Features are designed to assist - not replace - professional judgment and instructional oversight.
2. Core principles
- Educational purpose first. AI Features must be used only for legitimate educational or school-operational purposes.
- Human-in-the-loop review. Educators and authorized school personnel remain responsible for reviewing outputs before using them with Students or relying on them for school decisions.
- Privacy by design and data minimization. Think Tech seeks to limit the amount of Personal Information and Student Data used in AI workflows to what is necessary for the intended feature.
- No training on Student PII. Think Tech does not use Student personally identifiable information submitted through the Services to train public or generalized AI models.
- De-identification where feasible. Where the feature allows, Think Tech seeks to mask or replace student and teacher names before sending prompts to an external model provider.
- Security and contractual controls. Third-party AI providers used by Think Tech should be bound by written restrictions on retention, onward disclosure, unauthorized model training, and security.
- Testing for safety and quality. Think Tech seeks to evaluate AI Features for reliability, harmful content, and inappropriate or biased outputs before and during deployment.
- Transparency. Think Tech seeks to describe in plain language when AI Features are used and what role the Educator or user should play in reviewing outputs.
3. Acceptable uses
- helping Educators create or adapt lesson materials, prompts, rubrics, examples, and instructional supports;
- generating educator-facing suggestions, summaries, explanations, or draft content;
- supporting administrative efficiency, including categorization, summarization, or workflow assistance for authorized users;
- providing analysis or recommendations that remain subject to educator or district review and judgment; and
- improving service quality, functionality, performance, or safety using aggregated and de-identified information.
4. Prohibited or restricted uses
- using AI Features to make fully automated high-impact decisions about a Student - such as discipline, grade assignment, promotion, retention, placement, eligibility, or safety determinations - without meaningful human review;
- using Student Data for advertising, unrelated commercial profiling, or any purpose outside the district-authorized educational function;
- inputting more Student PII, special education information, health data, or other sensitive data than is necessary for the intended feature or school-authorized purpose;
- generating or distributing content that is harassing, discriminatory, defamatory, sexually explicit, violent beyond legitimate educational need, or otherwise inappropriate for an educational environment;
- circumventing academic-integrity rules, copyright obligations, or acceptable-use rules of the Educational Institution; or
- sending Customer Data to an external model provider that is not contractually bound to materially similar privacy and security obligations.
5. Data handling and provider expectations
AI prompts and responses may be stored inside Think Tech systems to return results, troubleshoot issues, and maintain service quality.
Where the current feature supports it, Think Tech replaces real student names with placeholders such as Student_001 before prompts are sent to an external provider; reverse-mapping metadata remains internal to Think Tech.
Third-party model providers may include OpenAI, Anthropic, and Google Cloud-based AI tooling as configured for a specific feature release.
Think Tech seeks enterprise-grade terms with AI providers that prohibit the use of Customer Data to train public or generalized models and that limit retention to the shortest feasible period compatible with the requested service.
Where contractually required by a district, Think Tech seeks to keep AI processing for U.S. district data within the United States.
6. Accuracy, bias, and educator responsibility
AI-generated outputs may be inaccurate, incomplete, outdated, biased, or inappropriate for a specific classroom, student, or district context. Educators and Educational Institutions remain responsible for reviewing outputs before relying on them in instruction, communication, or decision-making. Think Tech does not represent that AI-generated output is infallible or appropriate for every educational context.
7. Questions and reporting concerns
Security or safety concerns involving AI features, general AI, privacy, or policy questions: hello@thinktechsolutions.org