Since 2018, the United States has imposed a series of sanctions on Huawei under the banner of "national security," most notably by placing the company on the U.S. Commerce Department’s Entity List and subjecting it to export controls.

Where the pressure actually began

In 2019, Google complied with U.S. government requirements and restricted Huawei’s access to parts of Android.

That distinction matters, because what people casually call "Android" is really made up of two different layers: AOSP and GMS.

AOSP, the Android Open Source Project, is the open-source core of Android itself. Its code is publicly available, continuously updated, and free for anyone to use, modify, and commercialize. Huawei’s EMUI, Xiaomi’s MIUI, and the many Android variants people use every day are all built on top of AOSP. In earlier years, Chinese manufacturers often labeled these systems as some kind of "UI" as a modest way of saying they were customized versions of Android. Some brands later used the "OS" label instead, such as vivo’s OriginOS, but that naming difference does not imply a fundamentally different technical category.

GMS, or Google Mobile Services, is Google’s commercial service layer for Android: account systems, app distribution, backup, and other parts of the Google ecosystem. Outside mainland China, most Android phones ship with GMS preinstalled. In mainland China, some devices may include it in theory, but because Google services are blocked there, GMS is effectively unusable. Domestic phone makers therefore provide their own equivalents, such as Huawei’s cloud services and app store.

Google’s restriction on Huawei was mainly about GMS. Huawei could no longer continue shipping new devices with Google’s service layer, but it could still use AOSP. In reality, neither the U.S. government nor Google had the power to stop Huawei from using AOSP within the limits of open-source licensing.

For users in mainland China, this changed very little. Huawei’s China-market EMUI had never shipped with GMS in the first place, and Google services were already absent from everyday use there. So the idea that Huawei had its operating system supply "cut off" or had its throat "choked" by the loss of Android was always misleading in the domestic context.

Overseas, however, the situation was very different. Users there depended heavily on Google services, and losing GMS severely damaged Huawei’s phone sales in international markets. Huawei responded by pushing HMS and AppGallery as replacements abroad—essentially exporting the kind of service layer it had already built for China.

The birth of “Hongmeng”

Later in 2019, Huawei introduced a new operating system under the name "Hongmeng," or HarmonyOS.

The official framing was ambitious: unlike Android/Linux and iOS, this was presented as a brand-new microkernel-based distributed operating system for all scenarios, intended first for devices such as smartwatches, smart screens, in-car systems, and smart speakers.

But what should have been a practical developer-facing launch looked more like a concept reveal. There was no meaningful system demo, no substantial documentation, no example code—only presentation slides describing how advanced and disruptive HarmonyOS was supposed to be.

That vagueness was not accidental. The word "HarmonyOS" was being used as a broad container that could refer to more than one thing. Sometimes it meant OpenHarmony, and sometimes it meant HarmonyOS in the smartphone sense.

The system unveiled at that event was OpenHarmony, the open-source branch. It had nothing to do with Android. Its main foundation was Huawei’s previously released lightweight IoT operating system, LiteOS, which fits exactly the categories Huawei emphasized at the time: watches, displays, vehicle systems, speakers, and other smart devices. Was OpenHarmony developed by Huawei? Yes. Was it some unprecedented breakthrough? No. Systems like this are common in the IoT world.

It was not until 2021 that consumers actually saw HarmonyOS on phones. That version was effectively the continuation of EMUI under a new name. In other words, HarmonyOS for smartphones was still Android-based and remained closed-source. Huawei’s overseas Android-based devices even continued using the EMUI name rather than switching to HarmonyOS. Was this phone version independently developed from scratch by Huawei? Clearly not. And it was not some rare category either, because the wider smartphone industry already runs on Android.

From that alone, it should be clear that OpenHarmony and HarmonyOS were not the same thing. Their relationship was also nothing like the relationship between AOSP and GMS.

So is HarmonyOS open-source? Yes and no.

Is it based on Android? Yes and no.

Was it independently developed by Huawei? Yes and no.

That was the beginning of a long-running conceptual blur around the word "HarmonyOS." Different systems, different architectures, different development paths, different timelines, and different capabilities were bundled together under a single label. Intentionally or not, the result was a strong impression that Huawei had long possessed deep technical reserves and a far-ahead system architecture of its own.

A large amount of Chinese media coverage, including from technology outlets that should have known better, did little to clarify these distinctions. Instead, much of it abandoned technical precision in favor of unrestrained praise: "self-developed," "not just Android with a skin," and other slogans that further confused the public understanding of what independent R&D and technological innovation actually mean.

That kind of hype may have helped Huawei. It brought reputation and commercial advantage. But for the broader phone industry and for technology as a whole, it also risked crowding out the people doing harder, quieter, less glamorous work—the ones who might otherwise have earned the recognition.

Seen as a whole, the sequence is revealing: the loss of GMS did not materially affect Huawei’s domestic market, yet Huawei used that moment to promote a so-called new HarmonyOS to Chinese consumers. The entire episode often looked more like a marketing opportunity than a story of visible scientific or engineering breakthrough.

A naming strategy that spread across the industry

Huawei’s style of "innovation by renaming" also influenced the rest of the Chinese smartphone market. Competition there is intense, and once Huawei gained an advantage from the rhetoric of "independent controllability," rival brands were bound to look weaker on that front.

Afterward, others followed with their own operating-system branding campaigns. Xiaomi launched HyperOS. vivo launched BlueOS. A new round of system-level concept competition began.

HyperOS includes OpenVela for IoT, which roughly parallels OpenHarmony, while later versions of MIUI adopted the HyperOS name in a way that resembles Huawei’s HarmonyOS branding for phones. The difference is that Xiaomi explicitly stated HyperOS still uses Android as its foundation rather than implying it had somehow become something entirely separate.

BlueOS, at least for now, is also primarily aimed at the IoT side, again roughly mirroring OpenHarmony rather than Huawei’s smartphone system. vivo has said BlueOS will not be "compatible" with Android apps in the future. That wording is notable because Huawei popularized another fuzzy habit here as well: instead of plainly acknowledging that HarmonyOS on phones was Android-based, it was often described in a more mysterious way as merely "compatible" with Android apps.

After NEXT

By 2024, the HarmonyOS that Huawei had long framed as independently developed finally appeared in a form the company wanted to emphasize more strongly: HarmonyOS NEXT, also marketed as "pure-blood HarmonyOS."

From the public timeline, HarmonyOS NEXT appears to be the natural successor to HarmonyOS, as though Huawei had finally brought its long-promised system to maturity after years of development.

Architecturally, though, HarmonyOS NEXT and the earlier phone-based HarmonyOS are not the same lineage in any straightforward sense. The older HarmonyOS was an Android system built on AOSP. HarmonyOS NEXT is presented as a new system. By folding Android-based HarmonyOS and then HarmonyOS NEXT into the same overall "HarmonyOS" umbrella, Huawei created a seamless inheritance of the name even though the underlying systems differ substantially.

From a business perspective, Huawei used Android to build and retain a user base, then later tried to shift that user base into its own ecosystem. As a strategy made possible under open-source licensing, that is not inherently blameworthy.

It is also fair to say that HarmonyOS NEXT reflects more of Huawei’s own technical investment and does not use the Android kernel. But that does not automatically make it revolutionary. Compared with EMUI and the earlier HarmonyOS, HarmonyOS NEXT does not yet represent a disruptive leap in what users can actually do. For a long time, it is likely to remain immature, and it still has a very long road ahead before it can match the feature completeness and ecosystem depth of iOS and Android.

For users, the practical improvement may be limited, and in some respects the experience may even regress. Ecosystem separation will inevitably bring inconvenience. For app developers and software companies, the cost of adaptation rises. On the other hand, that also creates new work and new jobs, which is at least a benefit for employment.

Do we need a self-developed mobile OS?

Maybe China does need a truly self-developed mobile operating system.

Whether consumers actually need one is something the market will decide.

Huawei’s sanctions-era positioning gave it a kind of moral halo: standing up to pressure, resisting U.S. restrictions, carrying the banner of self-reliance. That image won broad sympathy and support. But the story is less heroic when viewed closely. Huawei benefited from Google’s open-source Android foundation for years, yet worked hard to distance itself from that relationship while claiming the achievements under the banner of "self-developed." In doing so, it also elevated itself above other Chinese Android phone makers that were relying on the same open-source base.

That makes the narrative feel less righteous than it first appears.