University of Washington: AI in Class Rejected by Parents

A Controversial Project at the University of Washington

The University of Washington recently considered conducting a study involving the use of cameras in kindergarten classrooms. The goal was to equip teachers with wearable cameras to record classroom interactions. These recordings would then be used to train artificial intelligence models. However, this initiative faced strong opposition from parents, leading to the project's cancellation.

In a document reviewed by 404 Media, researchers explained that the lead teacher could wear a camera filming their perspective, or a fixed camera could be installed in the classroom. These recordings were intended to capture normal interactions between teachers and students during regular school activities, automatically involving the children in the experiment unless their parents explicitly opted out.

Parents' Reactions and Project Cancellation

The parents' reaction was immediate and intense. They not only refused to give their consent but also vehemently contested the project. This opposition was strong enough for the University of Washington to decide to completely abandon the experiment. One parent, speaking anonymously to 404 Media, expressed concern about the use of their child's image in unknown AI tools and the potential risks of misuse.

Consent and Transparency Issues

The project was criticized for its lack of clarity and transparency. The documents provided to parents mentioned the use of videos to develop and evaluate AI models capable of analyzing the quality of classroom interactions. They also indicated that the data could be processed through cloud-based AI services, without specifying which companies would be involved or which models would be used.

Parents questioned the fate of the images of children whose participation they refused. Would they be blurred or deleted? How could authorized students be distinguished from others in a room full of young children in constant motion? One mother condemned the absence of true informed consent, highlighting that researchers promised to mask faces and names "to the extent possible," a reassurance deemed insufficient.

Language Barriers and Expert Criticism

The document, written solely in English, also posed a problem for many immigrant families who did not speak the language. No translated forms were available, further complicating understanding and informed consent.

Education specialists also criticized the project's lack of transparency. Faith Boninger, co-director of the National Education Policy Center, raised several unanswered questions: who would have access to the data, how long would it be retained, and who was actually funding this research?

A Worrying Trend in Education

The incident at the University of Washington is part of a broader trend where artificial intelligence is increasingly integrated into the educational field. Companies like OpenAI, Anthropic, and Microsoft are heavily investing in educational programs, and universities are multiplying partnerships to provide students with free access to generative AI tools.

However, this integration raises growing concerns among parents, who worry about the massive data collection necessary to train these models. The abandonment of the University of Washington's project reflects an increasing resistance to experiments perceived as intrusive.

In New York, a largely AI-automated high school project was recently abandoned after a mobilization of parents in front of City Hall. These events show that the acceptance of AI in education is not guaranteed, especially when it comes to children's privacy.