Microsoft has new AI privacy rules for schools
In response to growing concerns over student data protection, Microsoft has established a comprehensive set of legally binding safety and privacy standards for the deployment of artificial intelligence in classroom environments.
Developed in partnership with prominent labor organizations, specifically the American Federation of Teachers and its New York City affiliate, these guidelines allow school districts to seamlessly integrate strict protective rules into their technology contracts starting this fall. Under the newly negotiated terms, the technology giant commits to safeguarding sensitive user information by refusing to utilize student or educator data for training its machine learning models. Furthermore, the company has agreed to minimize overall data collection, completely prohibit the implementation of AI companion features, and mandate human oversight for high-stakes decision-making processes.
This proactive policy shift follows temporary bans on student-facing AI tools recently implemented by major public school systems in Los Angeles and New York City, as administrators sought time to construct appropriate guardrails. While many parents and educators remain highly skeptical of increased classroom screen time and automated tools, teachers' unions have emphasized the necessity of securing enforceable legal commitments directly from technology providers in the absence of robust federal regulation. To complement these defense measures, the labor groups are also collaborating with multiple leading artificial intelligence developers to finance training programs, ensuring that instructors are well-equipped to manage and integrate these rapidly evolving digital systems safely and effectively.
Summary generated September 10, 2026. AI summaries can make mistakes.
Read Original on The VergeCategory
Topic (AI-estimated)
AI & Machine Learning
85% confidence
AI Policy & Ethics
This category is an AI-estimated classification based on the article's content and may not be fully accurate.
Sentiment
Sentiment
Neutral
