The stage of the 2026 CCTV Spring Festival Gala served as a national-level showcase for China's AI industry. On one hand, dozens of humanoid robots from Unitree Technology flawlessly performed martial arts flips and synchronized dance moves, captivating a global audience overnight and drawing widespread acclaim from top international AI robotics experts. On the other hand, the comedy sketch "AI Partner" went viral across the internet, bringing the security risks associated with AI deepfakes into public awareness and making "AI trustworthiness and safety" the most prominent technology topic at the start of the year. These two contrasting phenomena not only demonstrated the vibrant vitality and innovative heights of China's AI technology to the world but also confronted a core challenge of the AI era: how to ensure technology possesses limitless potential while remaining safe, controllable, reliable, and trustworthy.
The Gala stage presented a dual reflection of Chinese AI, moving from a dazzling debut to a critical safety discussion. Commenting on the impressive performance of the Unitree robots, Agility Robotics research scientist Chris Paxton and Tesla Optimus machine learning engineer Yi Li offered insightful professional assessments. Yi Li stated that hardly any companies are now developing open-source humanoid robots based on non-Unitree architectures, as Unitree has completely dominated the open-source ecosystem in this field. Both experts concurred that the strategy of providing a fundamental development platform for the entire industry, which once made Apple successful, is destined to achieve the same for Unitree in the future.
This industry recognition was not achieved overnight. For the past five years or more, in the open-source robotics track largely overlooked by tech giants, Unitree has consistently focused on deep cultivation. It has continually supplied hardware-reliable, cost-effective, and fully open-source software-based robot dogs and humanoid robots to researchers worldwide in academia and industry. While this path may not yield massive short-term commercial profits, it has allowed Unitree to quietly build an ecosystem barrier. Today, nearly all cutting-edge research in global humanoid robot motion control is conducted on the Unitree G1 platform. Unitree may not have initially set out to build a global robotics platform empire deliberately. By focusing on making hardware extremely reliable and software exceptionally developer-friendly, it ultimately reaped the unexpected reward of becoming an "open-source industry leader" in a niche area neglected by giants, confirming the enduring truth in the tech world: innovation always happens at the edges.
Sharing the stage with Unitree's robots were other "visible AI" applications, such as AI digital humans and creative programs generated by large models. These used stunning visual effects to immerse hundreds of millions of viewers in a technologically infused festive atmosphere. Conversely, the viral sketch "AI Partner" showed the public another side of AI technology: the security risks posed by AI deepfakes, which can easily fabricate audio/video and generate false information, are no longer a distant industry concern but a real challenge relevant to everyone. Catalyzed by this national event, the market collectively realized that the AI governance and content security sector has transitioned from an optional add-on to a necessity for industrial development, poised for acceleration driven by both regulatory requirements and technological risks.
Beyond the stage, as millions marveled at AI's possibilities and grew wary of its potential dangers, an "invisible AI" force has been working diligently for over a decade in the field of trustworthy AI. This force, HAIZHI TECH GP (02706), safeguards the warmth of family reunions during the Spring Festival and builds a fundamental technical barrier for AI security at its core.
Operating behind the scenes, HAIZHI TECH GP uses trustworthy AI to strengthen the foundation of public livelihood during the Spring Festival reunion period. The smooth operation of the national transportation network under the pressure of billions of passenger trips during the Spring Festival travel rush and the stable electricity supply to households amidst peak holiday demand—this underlying sense of security and smooth functioning in people's lives—relies on the trustworthy AI foundation built by HAIZHI TECH GP using its graph-model fusion technology. Its chosen path bears a striking resemblance to Unitree's: eschewing the crowded, competitive arena of general-purpose large models, it selected the less-traveled road of trustworthy AI. It has focused on honing underlying technologies that address core AI challenges within vertical industries like energy, transportation, finance, and government affairs, sectors where giants have not deeply penetrated.
The two core aspects of public livelihood during the Spring Festival are "a safe journey home" and "a well-lit home." These critical industrial scenarios are precisely where the hallucination problems of general-purpose large models are unacceptable—a single erroneous dispatch could delay countless homecomings or disrupt family reunions. HAIZHI TECH GP's proprietary graph-model fusion technology fundamentally solves this core problem. It natively integrates knowledge graphs, often called the AI "digital brain," with large models throughout the entire process, rather than applying simple external patches for error correction. This ensures that every AI decision and output is backed by verifiable factual evidence and a traceable logical chain, perfectly meeting the "zero-error-tolerance" requirements of fields like transportation and energy. It also represents a core technical pathway to countering AI deepfakes and false information generation.
In the context of Spring Festival transportation support, HAIZHI TECH GP's technology is no longer a laboratory concept but is deployed across the entire chain of traffic management scenarios. Serving as the "smart core" for Shanghai's road transport, the company participates in building and maintaining the Yunlu Center for the Shanghai Road Transport Development Center, which handles core functions like city-wide road network monitoring and intelligent dispatch. During the travel rush, facing surging passenger volume in the Yangtze River Delta and doubled road network loads, this system operates 24/7, continuously capturing real-time traffic flow changes and accurately predicting congestion points. It provides comprehensive, gap-free decision support for Shanghai's travel调度, helping millions of passengers avoid congestion and travel smoothly.
Beyond smart dispatch for urban arteries, HAIZHI TECH GP extends its technology to the "last mile" of the journey home. It has undertaken upgrades for road facility management, completing full-chain enhancements for traffic sign management, road safety facility control, and rural road data support systems. This system can monitor the status of road safety facilities in real time and precisely support the maintenance and management of provincial highways and rural roads, ensuring that even remote village routes remain smooth and well-marked.
On the critical railway networks during the travel rush, HAIZHI TECH GP also acts as an invisible "guardian." Since 2024, it has successfully bid for projects including the upgrade of the Shanghai Railway Public Security Bureau's mobile police work smart policing system and the data modeling platform, achieving full front-end and back-end optimization of the policing system and enhanced data processing capabilities. During peak travel periods, this upgraded smart system can rapidly conduct information verification and handle police situations, significantly improving policing efficiency and strengthening the security防线 for railway travel.
On the front line of ensuring stable power supply during the Spring Festival, HAIZHI TECH GP's technology is deeply embedded in the entire process of smart grid operations. Using graph-model fusion technology, it maps the relational connections between equipment across the entire chain of power generation, transmission, transformation, and distribution. It analyzes trillions of grid operation data points in real time, accurately predicting potential failures in power equipment and optimizing power dispatch and consumption, thereby maintaining the high efficiency and stability of China's power grid under extreme holiday loads.
Apart from safeguarding public livelihood scenarios, HAIZHI TECH GP's technology also serves as a "safety lock" for industrial-scale AI implementation, directly addressing the security vulnerabilities highlighted by the "AI Partner" sketch. In financial fraud prevention, it can accurately identify AI-forged identities and fraudulent transactions through association graphs. In government regulatory scenarios, it can verify the authenticity of AI-generated content through logical reasoning. In enterprise digitalization, it ensures that every business decision made by an AI agent is traceable and accountable.
As AI application trends diverge, leading stocks are entering a favorable configuration window. If the Unitree robots on the Gala stage represent China's AI "technological appeal" to the world, showcasing the hardware prowess of China's robotics industry through visible marvels, then the "invisible AI" deeply cultivated by HAIZHI TECH GP in energy, transportation, and security sectors represents China's AI "technological substance"—safeguarding public livelihood and empowering industries. Their core philosophies are highly aligned: both have avoided crowded hotspots favored by tech giants, choosing instead to focus deeply on niche areas. They hone underlying technology with a long-term perspective, together forming a complete and robust ecosystem for China's AI industry. This outlines the most valuable characteristic of Chinese technological innovation: not chasing short-term hype or profits, but innovating at the edges and strengthening the foundation, ultimately ensuring that technology truly serves humanity, aiming for the stars while steadfastly guarding the warmth of countless homes.
Strategy analysts note that the AI application trend is far from over, and market adjustments are creating a golden window for configuring high-quality stocks. Subsequent market movements are expected to increasingly focus on leading companies with clear business models and rapid industrial implementation. As the dominant leader in China's AI hallucination elimination track, HAIZHI TECH GP has built a solid technical barrier with its graph-model fusion technology. It has achieved scalable, verifiable commercial deployment in core sectors like transportation, energy, and government affairs, perfectly aligning with the core investment theme of the current AI application wave. Coupled with its unique industrial value and strong earnings growth momentum, the company's stock possesses significant potential for upward movement in the future.