China’s PISA results show strong performance in four jurisdictions, while Beijing is introducing AI instruction in schools. California is developing its own approach.
The latest international education results have put four Chinese jurisdictions at the center of attention.
Students in Beijing, Shanghai, Jiangsu and Zhejiang — collectively identified by PISA as B-S-J-Z (China) — scored an average 612 in mathematics and 597 in science on PISA 2025. The OECD averages were 463 and 482.
But there is an important distinction: B-S-J-Z is not a score for all of China. It represents four participating Chinese jurisdictions.
And PISA can tell us how students performed. It cannot, by itself, prove that a particular curriculum, school schedule, tutoring system or teaching method caused those results.
What makes the Chinese example especially interesting for California is what is happening alongside those test results.
Beginning with the 2025–26 school year, Beijing required primary and secondary schools to provide students with at least eight class hours of AI education each year.
The city’s approach is designed to become more advanced as students progress through school, with younger students introduced to basic concepts and older students moving toward applications and more sophisticated AI-related learning. The guidance also incorporates AI ethics.
The significance is not simply that students are being given access to AI.
They are being taught about the technology.
That raises a question California parents and educators are already beginning to confront:
What should children learn before AI becomes an everyday tool for school, work and life?
California is not starting from zero.
The California Department of Education has developed “Learning With AI, Learning About AI,” statewide guidance addressing how schools can use and teach about artificial intelligence.
The state has also created an AI in Education Working Group under SB 1288 to develop guidance and model policy around the safe, equitable and effective use of AI in schools.
That makes this more than an international education story.
It is a California education story.
California is trying to determine how students can benefit from AI while protecting privacy, equity, critical thinking and the role of teachers.
That tension is already visible in Los Angeles.
Parriva recently examined the question in “LAUSD Students Can’t Use AI at School. What Happens When They Enter an AI-Driven Job Market?”, which looked at the gap between restrictions on student generative-AI use and the growing demand for AI-related skills in California’s workforce.
The issue is therefore not simply whether students should be allowed to use ChatGPT.
It is whether they are being taught how to understand the technology.
AI literacy is not the same as unlimited AI use
Students still need to learn how to read, write, calculate, research and solve problems without allowing a machine to do the work for them.
At the same time, they need to understand a technology that is rapidly becoming part of everyday life.
Those goals do not necessarily conflict.
A student can learn:
- how AI systems generate answers;
- how to recognize an unreliable or fabricated response;
- how to verify information;
- how AI can reproduce bias;
- how to protect personal information;
- when AI assistance should be disclosed; and
- how AI is changing different careers.
In other words, students can learn about AI without allowing AI to replace learning.
For Los Angeles parents, that distinction is becoming increasingly practical. Parriva’s recent guide to LAUSD’s ChatGPT and AI rules explains the district’s rules and why the educational purpose of using AI matters as much as the technology itself.
California has another problem to solve: equity
The AI question also intersects with an older California education problem.
Access to technology does not automatically mean equal access to high-quality education.
California’s Latino students continue to face significant educational disparities tied to factors including income, language support and differences in school resources, according to research examined in Parriva’s California education equity report.
That makes AI education an emerging equity question.
If AI literacy becomes an important academic and workplace skill, California will need to consider whether students across different communities are receiving comparable opportunities to learn how the technology works and how to use it responsibly.
There is not enough evidence to say an AI education gap already exists.
But the question is becoming harder to ignore.
The lesson for California isn’t “be more like China”
That would be too simple.
China’s PISA results provide evidence of strong performance in four participating jurisdictions. Beijing’s AI curriculum provides one example of how a school system is responding to artificial intelligence. Neither provides California with a ready-made blueprint. OECD PISA 2025 results
California has its own education system, priorities and challenges.
In fact, Los Angeles is simultaneously trying to reduce some student screen use while figuring out how students should encounter AI. Parriva’s recent look at LAUSD’s new screen-time rules shows why the issue is more complicated than simply adding more technology to classrooms.
The real question is what happens when those two goals meet:
How do schools protect the foundational skills students need while preparing them for a world in which AI will increasingly perform some of the work?
PISA cannot answer that question.
Neither can China’s curriculum.
But together, they give California a useful reason to ask it now.
The future of education may not be about choosing between traditional learning and artificial intelligence.
It may be about making sure students learn enough themselves to know when to use AI, when to question it and when not to use it at all.
