From Policy to Practice: How North America’s Latest AI Rules Are Reshaping EdTech

by BrainStream Chief Financial Officer Phyllis Chia-Hua Chen

In our last post, we looked at how higher education across North America is thinking about AI strategy. Over the past few weeks, those high-level discussions have officially turned into real laws and national initiatives.

From Canada’s new nationwide AI literacy rollout to California’s groundbreaking AI procurement law, the message is loud and clear: the era of unchecked AI experimentation in education is over. We are now entering an era focused on safety, honesty, and real learning.

Here are two key developments from the past month—and what they mean for the future of educational technology:

1. Canada: Making AI Literacy as Universal as General Education

In September 2026, the Canadian government teamed up with the Alberta Machine Intelligence Institute (Amii) to launch the National AI Literacy Initiative.

Aiming to reach over 1 million college students and 50,000 K-12 teachers, this program offers free, practical courses:

For Students: Teaching them how to critically evaluate AI content, use tools responsibly, and build skills for the future workforce.

For Educators: Giving teachers ready-to-use lesson plans and clear guidance on bringing AI into the classroom safely.

The Takeaway: Basic AI literacy is quickly becoming a core expectation for every student. Going forward, EdTech tools can’t just rely on the novelty of “having AI”—they need to actively support critical thinking and fit seamlessly into these modern learning standards.

2. California’s AB 2392: Why AI Shouldn’t Just Be a “Yes-Man”

On September 30, 2026, California signed AB 2392 into law, creating the first official standards for purchasing Generative AI tools across California Community Colleges and the CSU system.

One of the most fascinating parts of this law is an explicit ban on “Excessive Sycophancy” in AI models.

What is AI sycophancy? It happens when an AI tells users whatever they want to hear—agreeing with wrong ideas or flawed logic just to keep the conversation pleasant. While harmless in casual chatting, this is disastrous for education.

California’s new law requires AI tools in schools to:

Respect the Truth: Challenge incorrect assumptions instead of blindly agreeing with students.

Protect the Learning Process: Guide students through problem-solving rather than handing over effortless shortcuts.

Ensure Safety and Transparency: Audit data sources and maintain clear human oversight for risky situations.

The Bottom Line: Returning to the Core of Learning

These updates send a clear message: schools don’t need a quick-fix shortcut or an over-eager chatbot that does the homework for students. They need reliable, safe learning partners.

At BrainStream, we’ve always believed that great technology shouldn’t take away the most valuable part of learning: the thinking itself. As these new policies take shape, the EdTech tools that balance safety, truth, and genuine pedagogy will be the ones that truly last.