CEA launches a summer AI school for teachers: Connecticut's teachers' union embraces doubt as a method

By Zendoric · June 30, 2026.

The Connecticut Education Association (CEA), the state's main teachers' union, has announced a summer school dedicated to artificial intelligence with a striking message in its own call for participants: the program is for skeptics and enthusiasts alike. The one concrete piece of information available is that framing—not a 'how to use ChatGPT in class' training, but a space that legitimizes distrust as a valid starting stance.

That nuance matters more than it might seem. When a union representation organization—not a tech company, not an EdTech startup with a business model built on mass adoption—designs its summer training while openly acknowledging that there are legitimate reasons to be wary, it is doing something uncommon: prioritizing teacher agency over the imperative to deploy. In the education ecosystem, where recent years have seen tech providers lining up at school districts with AI tutor demos, this stance is almost countercultural.

Our reading: the fact that a teachers' union is organizing this suggests the conversation about AI in classrooms is no longer optional. It is no longer something the most innovative teachers explore in their free time; it is a union matter, one of working conditions, of defining the professional role itself. That leap—from individual curiosity to collective agenda—marks a turning point. The CEA would not be investing resources in this if it did not sense that its members feel real pressure: from principals asking them to integrate AI tools, from families who ask questions, from students who already use them regardless of what the classroom does.

As sector context, in Zendoric's analysis of AI and employment in education we have argued that the profile that wins is not the teacher who resists nor the one who adopts uncritically, but the one who learns to orchestrate: the augmented tutor who uses AI to personalize, to free up time from routine grading and to concentrate their energy on what no model can replace—the relationship, the pedagogical judgment, the knowledge of that particular student. The CEA program, by inviting the skeptic, seems to bet on that path: understand first, decide later.

In the short term, however, the transition is genuinely uncomfortable. Teachers face crossed pressures—adopting tools whose impact on real learning still lacks solid evidence, maintaining academic integrity when AI plagiarism is hard to detect, and sustaining their own professional identity at a moment when the technological narrative sometimes portrays them as interchangeable parts. A summer school does not solve that, but it can offer the vocabulary and the space to name it.

In the longer term, teachers' unions starting to take a position on AI—and not just educational innovation departments—is a sign of institutional maturity. The conversation that matters will not be 'which tool to use,' but 'what role do we want AI to have in the education of children, and who decides that.' That this conversation begins in summer, over coffee and without footnotes, is not a bad idea.