Imagine higher education AI as a single tree. In the first half of 2026, six distinct forces are watering it simultaneously. The interim results announcement from ABLE DIGITAL (02687) shows on-hand orders reaching RMB 505 million by the end of the reporting period, a 34.7% year-on-year increase that outpaces the 19.1% revenue growth seen in the same period.

Orders serve as a leading indicator of demand, and this round of demand expansion is not driven by a single policy stimulus or short-term event. Rather, it is a structural resonance where six forces converge: national strategy, industrial transformation, teaching practices, learning methods, research paradigms, and the large model ecosystem.

The nation's need: infrastructure for an independent knowledge system. During the reporting period, the state issued a concentrated wave of major science and education policy. The "Education Powerhouse Construction Planning Outline" proposed building China's independent knowledge system with integrated development of education, technology, and talent, further clarifying the strategic position of knowledge graphs as knowledge system infrastructure. The "AI+ Education Action Opinions" and "Action Plan," along with the "15th Five-Year Plan," were released intensively, accelerating the inclusion of intelligent teaching and virtual experiments into new educational infrastructure. The "Opinions on Accelerating Educational Digitalization" promote deep integration of large models with education and teaching, while the "101 Plan" for basic disciplines continues to expand into new engineering and new agriculture fields. These policies form a systematic deployment centered on building an independent knowledge system, university research breakthroughs, and innovative talent cultivation.

National-level recognition has already translated into substantial positioning. In May 2026, the Ministry of Education released the first batch of 18 national-level higher education intelligent agent directories at the World Digital Education Conference, with projects co-built by the company included. Within the "101 Plan" core project, a national higher education reform initiative led by academicians from both academies, the company has deeply participated in co-building efforts covering computer science, physics, basic medicine, economics, chemistry, mechanics, atmospheric science, agricultural engineering, smart agriculture, geology, intelligent medical engineering, pharmacy, Chinese medicine, and philosophy. The company is extending from a technology supplier to a standard co-builder.

Industry's need: AI talent supply and industry-education integration. Industry-side demand comes from two directions: first, the large-scale talent gap created by AI technology penetrating various sectors, and second, the urgent need for composite talent in traditional industries undergoing digital transformation. The company's business has broken through campus boundaries, extending into high-value verticals such as healthcare, scientific research, and industry. Its cooperation partners cumulatively span multiple hospitals, academic institutions, research organizations, and leading advanced manufacturing enterprises. On the international front, the China-Indonesia AI Talent Factory (AITF) project co-built by the company officially signed a cooperation agreement in Jakarta in August 2026. This project is among the first batch of national digital talent training pilot programs under Indonesia's "National AI Strategy 2020-2045," aiming to cultivate 1,000 skilled AI professionals between 2026 and 2029. Additionally, at the Third SCO Green Energy Academicians Forum, the company inaugurated the China-Kyrgyzstan Clean Energy Storage Joint Laboratory, applying its proprietary AI technology to cross-border platform construction.

Teachers' need: evolving from tools to intelligent collaboration partners. The core need of faculty is to be freed from repetitive teaching tasks so they can focus on instructional design and academic innovation. The company's AI agents adopt a dual-layer architecture of "course agents" and "school agents." The course layer focuses on intelligent lesson preparation, adaptive question generation, and learning analytics for full-process intelligent teaching, while the school layer handles local deployment and data security. The Panorama Space provides teachers with a physical offline carrier through immersive displays and multimodal human-computer interaction terminals. During the reporting period, AI agent contract signings grew 62.3% year-on-year, and Panorama Space customer numbers grew 77.8%, directly reflecting the acceleration of demand from the teaching side.

Learners' need: personalized and immersive learning experiences. What learners need is a personalized path tailored to their individual abilities, not standardized knowledge indoctrination. Based on the "Da Ming Bai" large model and knowledge graph system, the company provides core capabilities such as personalized learning path recommendations, knowledge reasoning Q&A, and retrieval. Physics AI virtual experiments expand knowledge from two-dimensional semantics to three-dimensional physical space, integrating spatial relationships, operational logic, and real-time feedback, offering learners repeatable practice environments for high-risk, high-cost, and high-complexity experiments. During the reporting period, physics AI virtual experiment contract signings grew 52.6% and customer numbers grew 46.4%, with applications broadly penetrating cutting-edge engineering fields such as unmanned systems and smart logistics.

Researchers' need: a disciplinary foundation for AI for Science. Research paradigms are being reshaped by AI. The company has built a product matrix around C9 League universities covering medicine, mechanics, nuclear technology, marine intelligence and unmanned systems, atmospheric science, and atomic nuclear physics. The Peking University Medicine Future Learning Center is a landmark collaboration. These disciplinary large models and knowledge graphs provide researchers with end-to-end AI assistance spanning literature review, knowledge reasoning, and virtual experiments. The company explicitly uses "AI for Science" as its official standard formulation, only applying it when collaboration content has clear research attributes, reflecting a serious positioning toward research scenarios.

The need of large models: vertical data and scenario landing. General-purpose large model vendors urgently need high-quality vertical industry data and scalable implementation scenarios, which are exactly the core assets ABLE DIGITAL has accumulated over nearly two decades. During the reporting period, the company signed strategic cooperation framework agreements with Alibaba Cloud (June), Volcano Engine (July), and Baidu Intelligent Cloud (August). With Alibaba Cloud, it jointly built the multi-agent platform "Da Ming Bai Polymas" based on the open-source AgentScope Java framework. With Volcano Engine, collaboration focuses on four areas: large model knowledge empowerment, virtual-real integration training, AI audio-video interactive teaching, and digital talent cultivation. With Baidu Intelligent Cloud, it integrates Qianfan large models, ERNIE series models, and professional encyclopedia knowledge resources to jointly develop vertical intelligent agents for education, industry, and research scenarios. The fact that three leading general-purpose large model vendors chose to cooperate with the company rather than replace it within two months essentially reflects consensus on the collaborative path of "general foundation + vertical expertise + campus scenario landing."

These six forces do not exist in isolation but reinforce and amplify one another. National policy provides direction and market space, industrial demand supplies capital and application scenarios, teachers and learners form the ultimate user base, researchers expand the knowledge frontier, and the large model ecosystem offers technological leverage. All six forces converge on the same tree: structured knowledge assets as roots, multimodal AI as the trunk, a nationwide delivery network as branches, and multidisciplinary scenarios as leaves. During the reporting period, revenue grew 19.1%, gross profit grew 28.0%, on-hand orders grew 34.7%, and high-value customer numbers grew 26.7%. This set of data represents the concentrated realization of the six-dimensional forces transitioning from policy vision to commercial execution. When a tree's roots have run deep for two decades, and six forces happen to water it simultaneously, its growth is not a passing breeze but a structural inevitability.