开放标准下的封闭陷阱：SGP.32 蜂窝物联网锁定争议浮出水面

Markdown Content: The telecom industry understands the value of open standards. It is why the GSMA spent years developing SGP.32 – the new eSIM IoT specification designed to make cellular connectivity more flexible, portable, and scalable for connected devices. Open standards create healthy competitive markets. They allow customers to choose between vendors, encourage innovation, reduce dependency on any single supplier, and enable interoperability between products and services. Cloud computing was built on open Internet technologies, yet many organizations now struggle with cloud lock-in. Open APIs have not prevented platform dependency in areas ranging from social media to enterprise software. Even standards-based cybersecurity frameworks can fragment when different trust models emerge around the same underlying standard. In theory, the promise of SGP.32 is straightforward. A fleet of connected devices should be able to change connectivity providers throughout their operational life without requiring physical intervention. Connectivity becomes a flexible software decision rather than a one-time hardware decision. The concern is that some implementations of SGP.32 are introducing new forms of operational and cryptographic lock-in that, while fully permissible within the current standard, may significantly reduce the freedom many customers assume the standard provides. It is the difference between compatibility and interoperability. Compatibility determines what a device can do today. Interoperability determines how much freedom its owner still has 10-15 years from now. If its connectivity architecture cannot adapt to local coverage variations, commercial realities, regulatory requirements, or future network changes, expansion can quickly become difficult, expensive, or commercially unviable. In parallel, the continued evolution of mobile networks, including the retirement of 2G and 3G, the eventual transition away from 4G, and shifting operator strategies around NB-IoT and LTE-M, all point to the same reality: no one can accurately predict the connectivity landscape 10 to 15 years from now. That makes preserving flexibility and freedom of choice more important than ever. The original eIM effectively becomes the gatekeeper to every future connectivity decision. If a customer cannot change the eIM, they may be unable to access entirely new ecosystems of connectivity providers that become necessary over the lifetime of their devices. The risk is not that a device stops working. The risk is that it cannot function reliably or remain commercially viable in a new market because of a technical decision made years earlier that locked the OEM into a restrictive connectivity ecosystem. What should have been a straightforward software-level change instead becomes a costly and complex infrastructure – and potentially hardware – problem. In some cases, the economics may no longer stack up, resulting in delayed expansion plans or canceled rollouts altogether. A growing debate within the industry concerns the use of self-signed certificates rather than GSMA-recognized trust frameworks. In these scenarios, devices may technically be SGP.32-compatible while remaining dependent on a predefined group of providers that recognize those credentials. A solution can legitimately describe itself as SGP.32-compatible because it supports the functions defined by the standard – downloading profiles, switching profiles, and managing connectivity remotely. SGP.32 allows an eUICC – the SIM component that manages connectivity credentials – to be associated with multiple eSIM IoT Remote Managers (eIMs). It also provides a framework for remotely managing eSIM profiles on IoT devices using mutually authenticated channels between the device and the provisioning platforms. After years of work, the GSMA should take the SGP.32 standard one final step further and make interoperability a requirement for full certification, rather than an expectation left for the market to deliver voluntarily. No one expects to walk outside 10 or 15 years from now and find that cellular networks have disappeared. But no one can predict exactly how the market will change either. Operators will merge. Regulations will shift. Coverage footprints will evolve. Spectrum will be reallocated. New technologies will emerge. Others will disappear. The same applies when products expand internationally. Connectivity architectures that work perfectly in Europe or North America may not align with network availability, commercial requirements, or regulatory conditions elsewhere. That principle applies equally to connected devices. The question is no longer whether devices can connect. The question is whether the industry is building an ecosystem that will remain adaptable when today's assumptions inevitably change. Because they will. There is no more ubiquitous global connectivity network than cellular. The decisions being made around eSIM IoT today will influence how billions of connected devices are deployed, managed, and evolved over the coming decades.