When the AI wave surges from laboratories into the industrial deep waters in 2026, Chinese enterprises are standing at an unprecedented crossroads: embrace an open architecture to achieve autonomy and control, or remain trapped in closed-source "black boxes," enduring vendor lock-in and runaway costs?
Recently, at a SUSE media interview in Shanghai, Peter Lees, Vice President and Head of Solutions Architecture for SUSE Asia-Pacific, candidly stated in an interview with the National Business Daily that his previously raised viewpoint—"Asia-Pacific enterprises will face a life-or-death choice between complete rejection or embracing transformation in 2026"—holds equally true for the Chinese market.
In Peter Lees' view, the core anxiety for Chinese enterprises today is not whether to adopt AI, but how to choose a technological path that won't leave them "locked in." The short-term allure of closed-source platforms is evident: ready-to-use, rapid deployment, seemingly offering a one-click solution to all infrastructure challenges. However, the price is equally fatal: subsequent price hikes, ecosystem exclusivity, and limited iterative upgrades. Once bound, it becomes difficult to break free, potentially reducing enterprises from technology leaders to "vassals" of a single vendor. This pain point, amid the fierce battle between open-source and closed-source approaches in the enterprise market, has become a core challenge the entire industry must collectively confront.
The 2026 Turning Point: China's AI Anxiety Converges with Global Trends
SUSE is a leading global independent open-source software company. Its products and solutions serve hundreds of thousands of enterprises worldwide, with over 60% of Fortune 500 companies relying on its technology to build their core IT architectures. It maintains a leading position in enterprise Linux, container cloud, and edge management sectors. From this industry standing and technical foundation, Peter Lees observes that, against the backdrop of the large-scale rollout of global open-source models and the intensifying competition in the enterprise market, the uniqueness of the Chinese market lies in its more aggressive pace of transformation and demand for implementation.
According to Peter Lees, the rhythm of Chinese enterprises in AI development and practical application is noticeably faster than in other Asia-Pacific regions, with a greater emphasis on "quick results, autonomy/controllability, and data compliance." Companies are unwilling to compromise with black-box models, especially in heavily regulated industries like finance, healthcare, and manufacturing, where demands for data sovereignty and model interpretability are nearly stringent. Lees states that this demand aligns closely with SUSE's core proposition: adhere to open source, preserve future options, and reject vendor lock-in.
Peter Lees believes that what enterprises need is not a "single solution," but an open, compatible, and extensible foundation. Regardless of the upper-layer applications, models, or hardware adopted, the underlying infrastructure should maintain interoperability—imposing no restrictions or bundling—allowing enterprises to retain initiative throughout technological iterations.
In 2026, Agentic AI has become an industry buzzword, as AI evolves from passive response to proactive execution, placing unprecedented demands on the stability, lifecycle, and automation capabilities of infrastructure. At this juncture, SUSE launched SLES 16, defined as the world's first enterprise-grade Linux system designed for Agentic AI. Xianyang Su, Director of Solutions Architecture for SUSE Greater China, told reporters that this aims to directly address two core industry contradictions.
First, it resolves the balance between long-term stability and rapid innovation. In the interview, Su emphasized that SLES 16 offers "an official support lifecycle of up to 16 years," a design intended to align with the lifecycle needs of enterprises' core business systems. He explained that while AI technology iterates rapidly, changing every few months, enterprises' core systems and production operations cannot be frequently restructured accordingly. A stable 16-year foundation means enterprises can freely adapt to the rapid iteration of upper-layer AI models, applications, and hardware while keeping the base layer unchanged. It also enables proactive fixes for underlying issues, fundamentally solving the industry's dilemma of "innovating on shaky foundations."
Simultaneously, the built-in MCP (Model Context Protocol) toolkit further embeds AI capabilities into the operational bedrock. It acts as an intelligent assistant within the system, capable of automatically executing policy validation, performance monitoring, and anomaly troubleshooting, freeing IT teams from tedious, repetitive tasks. Where 70% of resources were once spent maintaining legacy systems, this can now be compressed to lower levels, allowing human and financial resources to be directed towards high-value innovation.
Second, it tackles the challenge of large-scale edge management in smart manufacturing. As Chinese manufacturing enters deeper waters, the unified deployment, upgrading, and maintenance of massive edge devices have become key bottlenecks constraining intelligent transformation. In traditional models, production line upgrades might require hours or even days of downtime, causing significant production losses. SUSE's solution is a lightweight containerized edge architecture based on Rancher and K3s. Through centralized management, edge devices can achieve remote, zero-downtime deployment, version switching, and one-click rollback.
Open Source as a Solution: Overcoming AI Misconceptions While Balancing Cost and Compliance
In the process of enterprises accelerating the deep integration of AI and smart manufacturing, behind the glamour of technological adoption, numerous common practical problems have also surfaced.
Many enterprises have fallen into dual misconceptions in cognition and execution during their transformation: either blindly following technological trends, or neglecting compliance and security bottom lines, or facing uncontrollable investment costs. This ultimately leads to hindered project progress and transformation outcomes falling short of expectations. Addressing these common industry pain points, the two SUSE executives deconstructed them one by one in the interview, providing clear paths forward based on their own technological philosophy and practical experience.
During the interview, the two executives pointed out three prevalent misconceptions in current enterprise AI transformation. The first misconception is pursuing AI for its own sake, lacking clear business objectives. Pressured by competition, many enterprise executives blindly initiate AI projects without a clear understanding of the business pain points or implementation paths, leading to misallocated resources and stalled projects.
Peter Lees emphasized that the essence of AI is problem-solving, not chasing trends. Enterprises must first define "what business metrics AI should improve" before choosing an open and flexible technology stack, avoiding paying the long-term price of being locked in for short-term convenience.
The second misconception is prioritizing innovation over compliance and security. "Many enterprises, in their AI transformation, focus solely on innovation while neglecting compliance and security, ultimately causing project failure," Xianyang Su added. This is a very common and fatal misconception in the AI transformation of Chinese enterprises, especially in heavily regulated industries where compliance and security are prerequisites for AI project implementation—if compliance requirements cannot be met, even the best projects cannot be launched.
The final misconception is the loss of control over AI investment costs, deterring small and medium-sized enterprises (SMEs). "High GPU computing power costs, low resource utilization, and uncontrolled invocations are the main reasons for soaring enterprise AI costs," Peter Lees admitted. This problem plagues not only SMEs but also troubles many large and medium-sized enterprises. Especially in 2026, as large-scale application of AI models increases demand for GPU computing power, its cost has become a "heavy burden" for enterprise AI transformation.
He noted that many enterprises, particularly SMEs, hesitate to start AI transformation precisely because they fear cost overruns—investing heavily in expensive equipment and specialized teams, only to find costs spiraling out of control, potentially leading the enterprise into financial difficulties.
SUSE's proposed solution is to help enterprises achieve cost control through open-source architecture and intelligent tools. Among these, the SUSE Observability tool is a core instrument. "This tool enables end-to-end monitoring of AI services, GPU utilization, and resource consumption, precisely identifying inefficient and wasteful areas," Peter Lees stated.
For external clients, this means quantifiable cost optimization. Xianyang Su added that many clients, after using this tool, have achieved improved computing resource utilization and cost reduction. "For instance, some manufacturing enterprises used this tool to monitor computing power consumption on edge devices, optimized resource allocation, and improved operational efficiency at the edge without additional hardware investment, thereby reducing operational costs. Some AI companies used it to monitor GPU usage, avoided computing power waste, and lowered costs."
Looking ahead to 2026, open source is no longer an option but a "required answer" for enterprise AI transformation.
For capital markets and industry participants, this signal is equally clear: enterprise IT investment in 2026 will shift from mere competition in AI applications to a re-evaluation of the value of the infrastructure foundation. Technology service providers with open-source DNA, long-term support capabilities, and synergistic edge and AI competencies will become core partners in enterprise digital transformation and will occupy more advantageous positions in the impending industry reshuffle.
