From September 8 to 11, 2026, UNESCO Digital Learning Week was held at UNESCO Headquarters in Paris, France. Centered on the theme “Education in the age of AI: Facts | Frictions | Frontiers,” the event brought together ministers of education, senior experts, representatives from the technology industry, educators, and students from around the world.
On the opening day, more than 25 ministers of education and their representatives issued a joint statement calling for education to continue being protected as a human right and a common good in the age of artificial intelligence. The statement emphasized that “it is of no benefit to learning if artificial intelligence replaces human relationships and human judgment.” As one of the most influential multilateral platforms in global education, this year’s Digital Learning Week focused on frontier issues such as the rise of agentic AI, AI sovereignty, student and teacher agency, and the scaling of personalized learning.
The conference gathered a high-level lineup of global participants, including education ministers, senior experts, and representatives from the technology sector. Among them were Chen Qun, Assistant Director-General for Education at UNESCO and former President of East China Normal University; Amandeep Singh Gill, United Nations Under-Secretary-General; H.E. Ms Sarah Al Amiri, Minister of Education of the United Arab Emirates; Sal Khan, Founder and CEO of Khan Academy; Åsa Regnér, Deputy Director-General of UNESCO; and H.E. Mr Rodolfo De Carvalo Cabral, Deputy Minister of Education of Brazil.
Squirrel Ai Co-founder and International CEO Dr. Joleen Liang was invited to attend the conference and delivered a keynote speech titled “From Precision Diagnosis to Scalable Impact: What AI Changes in Personalized Learning.”
This year’s theme aimed to examine the real impact of AI in education, the practical barriers it faces, and its future directions. Key topics included the rise of agentic AI, AI sovereignty, student and teacher agency, and the scaling of personalized learning. Within this high-level forum, Dr. Joleen Liang’s participation and remarks enabled the global education community to better understand, from the frontlines of large-scale implementation, how AI can more effectively serve human development.
From Precision Diagnosis to Scalable Impact: The Core Logic of AI-Driven Personalized Learning
In her keynote speech, Dr. Joleen Liang focused on “From Precision Diagnosis to Scalable Impact: How AI is Transforming Personalized Learning,” directly addressing the long-standing scalability challenges in personalized education and offering new approaches to solving them.
In traditional models, personalized learning often relies on teachers manually adjusting instructional content. While this approach may work in small-scale settings, it is difficult to sustain at scale. Squirrel Ai argues that the key to overcoming this challenge lies in enabling AI to deeply understand the learning process and achieve precise diagnosis.
To this end, Squirrel Ai has developed a multimodal adaptive education large model powered by a fine-grained knowledge graph and multimodal data analytics. It not only identifies gaps in a student’s knowledge, but also traces those gaps back to underlying cognitive weaknesses, thereby dynamically generating individualized learning pathways.
This leap, from identifying “what is wrong” to understanding “why it is wrong”, makes personalized education less dependent on individual teacher experience and more scalable and replicable. Squirrel Ai’s MCM (Mode of Thinking, Capacity, Methodology) training system embodies this framework.
As AI tutors and intelligent agents continue to emerge, Squirrel Ai’s practice demonstrates that the key to scalability is not simply increasing computing power or user numbers, but ensuring that every learner receives precise, continuous, and cognitively meaningful support.
Human–Machine Collaboration and Educational Sovereignty: Value Choices in the Path to Scale
In the panel discussion, Dr. Joleen Liang joined guests from organizations including the OECD, MIT, and Stanford University in a dialogue on “Why is personalized learning difficult to scale, and what difference does AI make?”
The discussion focused on how AI can enable more precise and scalable personalized learning, particularly in understanding learning processes and diagnosing individual learning gaps, as well as how these technological capabilities can be translated into tangible impact across different education systems.
AI takes on tasks such as diagnosis, personalized instruction, and repetitive analysis, freeing teachers from heavy administrative and routine workloads. This allows educators to focus on emotional engagement with students, value guidance, and judgment and intervention in complex learning contexts. Within this human–machine collaboration model, both AI-driven education and the role of teachers are being redefined: AI serves as a “personalized learning engine” that provides deep insights, while teachers evolve into “learning strategists” and “mentors for life development.”
Facing emerging frontiers such as “agentic AI” and “frugal AI,” Squirrel Ai remains committed to a vertical domain-specific large model approach in education. From Squirrel Ai’s perspective, education requires precise, high-value domain-specific data, and the generalization capabilities of general-purpose models are insufficient to meet this requirement. At a time when AI sovereignty and human agency have become global topics of discussion, this strategic choice reflects a deeper commitment to the public nature of education and human development.
As artificial intelligence reshapes the global education landscape, the true transformation does not lie in technology itself, but in whether it can ensure that every learner is seen, understood, and empowered. Technology is becoming a strategic partner in unlocking human potential, and the ultimate goal of scalable personalization is to enable each learner to grow at their own appropriate pace.
Squirrel Ai’s practice demonstrates that the deep integration of AI and education is not merely an improvement in efficiency, but a contemporary realization of the long-standing educational ideal of “teaching students according to their aptitude.” This sharing highlights Squirrel Ai’s exploration at the forefront of educational technology, and its sustained focus on human development also provides an important reference for how global education can responsibly integrate intelligent technologies.
