The headline news from Huawei’s analyst conference a few weeks back was the admission that, circa 2018, 5G may be somewhat overhyped. The larger story out of the event, however, was about Artificial Intelligence (AI). More specifically, it was about the AI opportunities Huawei hopes to tap and the obstacles it will face on the way to realising them.

If you’ve lived through the transition from 3G to 4G, the comments from Huawei’s rotating Chairman Eric Xu about 5G might not have struck you as anything less than simply honest: the first 5G release (3GPP release 15) is just a starting point, expectations around 5G are running too high, the differences between advanced 4G technologies and initial 5G deployments are minimal. It’s reminiscent of the early days of LTE when HSPA+ promised 4G-like speeds and the new networks being deployed promised faster mobile broadband and new business models that would follow.

The impetus for these comments? An analyst question about why 5G hadn’t factored more prominently in the keynotes delivered by execs on the opening morning of the event. Given the vendor’s 5G product launches, 5G messaging, and massive 5G presence at Mobile World Congress, it was a fair question. But it only arose because Huawei spent the morning talking up its plans around a completely different topic: AI.

To be fair, the overarching focus on AI didn’t go completely unnoticed or unreported. And, for anyone who has followed Huawei or the telecom and tech world for any amount of time, it was also a focus that was easy to predict. Put aside, for a moment, the fact that AI was a major component of Huawei’s MWC messaging and its pre-MWC briefings. Put aside the fact that AI factored into nearly every vendor’s MWC messaging along with the product launches that followed. Put aside the fact that AI is making its way into the widest array of products from telecom network infrastructure to smartphones to security platforms, IoT platforms, back office systems and everything in between. Beyond the fact that it’s an industry-wide hot-button, AI is a strategic initiative for China as a whole.

Where the People’s Republic may have missed out on driving the standards around globally important technologies in the past, it does not plan to miss out on the AI opportunity; Huawei (as an example of China’s technological know-how and success and a key supplier into Chinese operators and businesses) would need to fit into this. And where Huawei was looking for an asset or strategic focus to link its diverse businesses? A comprehensive, holistic thread calling out cross-company synergies? Whether applied to enterprise products, handsets or carrier networks, there’s a role for AI R&D to be leveraged across all of Huawei.

It’s a compelling message, no doubt. But it’s far from a slam dunk. Beyond the fact that competition around Artificial Intelligence solutions will run high (as it should around any important, new technology) Huawei has a handful of its own obstacles to navigate.

Coincidental Synergies: It might make sense to position AI as a common theme across different business units, but that doesn’t mean AI connects them in a meaningful way. Different R&D efforts. Different use cases. Different customers. Diverse businesses might all be focused on AI, but that doesn’t mean they are coordinated – and if they’re not, their efforts could be wasted or even conflicting.

Promises vs. Products: In discussing how it would realise its AI ambitions, particularly in the enterprise, Huawei sent a very clear message: watch this space. While an architectural vision is in place, new products were promised in tandem with its Huawei Connect event later in the year (scheduled for October in Shanghai). Time and again, the vendor has proven its ability to deliver on its promises. Where AI competition is running high and there’s no shortage of solutions in the market, it will need to do more telling and less showing if it wants to be taken seriously.

Chinese AI: In a world where governments are raising flags over the use of telecom networking gear and even phones from Huawei and ZTE, how do you think its AI products may be received? Before you answer, recognise that AI use cases include access to personal data and the manner in which AI systems make their decisions aren’t always visible; the biases and assumptions implicit in those solutions may well be unknowable.

If AI has the potential to transform consumer experiences, enterprise operations, and telecom networks, it’s only natural that the competitive stakes around it would run high – and that success for any would-be AI player is far from assured.

And there’s the rub; the obstacles Huawei faces in executing on its AI vision stretch well beyond Huawei. They could just as easily be applied to most of its competitors. The need to deliver market proven products and not just product visions; the value of executing on AI as a real synergy across product lines and not just a convenient marketing tool; the potential for any vendor to see their AI solutions and sales efforts get wrapped up in global politics.

If this last point seems like exaggeration, consider some basic AI realities. That it touches lots of data; personal data. That it has the potential to link diverse products and systems in ways other technologies have not. That it’s not always clear how AI systems make their decisions – meaning that there needs to be implicit trust in the systems’ purveyors. If political and trade tensions around telecom technology continue to rise, AI may be everyone’s opportunity to lose.

 – Peter Jarich, Head of GSMA Intelligence

The editorial views expressed in this article are solely those of the author and will not necessarily reflect the views of the GSMA, its Members or Associate Members.