BETA
This is a BETA experience. You may opt-out by clicking here

More From Forbes

Edit Story

Apple’s AI Plans In China—A Real Need To Tread Carefully

Following

Apple’s recent AI news has been a huge surprise—and now the latest rumor has reportedly been confirmed; whatever happens next, it looks like everything has changed for iPhone, with serious security and privacy risks to be navigated…

Apple is more private and secure than Android. That has been Apple’s mantra since the beginning, governing its approach. But those differences are diminishing, as Google ramps up its security and privacy and Apple makes concessions of its own.

Cue the latest news from China. According to local media, last week’s surprise report that Apple was discussing a generative AI deal with China’s Baidu has now been confirmed. The deal is done. Chinese generative AI is coming to iPhone—in China.

Baidu’s AI model is approved by a Chinese state with which the company has strong links. You might recall reports that even suggested military AI ties—albeit these were denied. Even so, this is interesting territory for privacy-focused Apple.

MORE FROM FORBESGoogle's New AI Search Goes Horribly Wrong-M Is For Malware

“Apple will use Baidu’s Ernie Bot for its iPhone 16, Mac OS and iOS 18 in China,” the South China Morning Star claimed, “according to a report on Monday by China Star Market, [and] follows a report on Friday by the Wall Street Journal that said Apple had held initial talks with search engine giant Baidu.”

The AI landscape in China is (unsurprisingly) regulated, with only government approved models allowed. And while the main focus of those approvals is censorship and the prevention of dissent, with China there is always the backdrop context of surveillance, monitoring and controls. This is an area Apple has navigated since its earliest days in China, and which its growth there has exacerbated.

But AI is the most difficult path to tread, because it has the potential to expand its role on our smartphones and in our lives so rapidly. And because granting or denying it access to our most sensitive information is going to be a continual challenge. Baidu is obviously subject to the same national security laws that are debated at such length when it comes to ByteDance/TikTok, Huawei and others.

When the speculation first started about Apple’s generative AI plans for iOS 18, the thinking was that everything would be on-device, that Apple would design hardware and software to ensure its privacy and security extended to the new world of AI. This would be a logical extension of its expanded end-to-end encryption and tight user controls. What happens on the iPhone, stays on the iPhone, as they would say.

There were reports that Apple was even testing its on-device AI results against Chat-GPT amongst others, to ensure results stacked up. But then came the news that the iPhone maker was talking to Google about Gemini and everything changed.

From a privacy standpoint, Google’s AI is cloud first, notwithstanding its Nano for lighter-weight on-device functionality. The idea that Apple was contemplating this Google technology at the heart of its iPhone surprised us all. It became the tech talking point of the year—at least until the DOJ stepped forwards.

MORE FROM FORBESYes, You Can Export Your WhatsApp Chats To Telegram-But Don't

But while it seemed an unlikely alliance, given Apple’s and Google’s different approaches to AI, cloud, encryption and privacy, it also had a ruthless commercial logic. Google is leading the way with generative AI and is streets ahead on applying this to smartphones and their apps and services. It has also paved the way for such a deal with the exclusive iPhone search arrangement. And time is of the essence.

This is the latest example of iPhone and Android becoming more closely aligned—iPhone’s expected mix and match approach, with Google Gemini in the cloud and its own AI on device matches Samsung’s own approach, which also seeks to balance AI features with user privacy and data security. Cue another parallel. Because Samsung has that exact same Gemini/Baidu combination.

Samsung’s answer is “a hybrid approach that combines on-device and cloud-based AI. Besides ensuring seamless usability, this lets users limit some features to function entirely on-device, giving them greater control over what they do with their data.”

This is exactly what Apple now seems to be contemplating for iPhone.

“When integrated into phones, AI is, to put it simply, a revolution,” Samsung’s TM Roh said ahead of this year’s MWC. “It’s also important to raise the standards of security and privacy in this new era of data-intensive mobile experiences.”

At the time Samsung was pushing out these messages, I assumed Apple would be more locked down as regards on-device versus cloud AI, with privacy considerations front and center. I also noted the balancing act Samsung had to play with both Google’s and its own AI on its devices. I did not expect Apple to be in the exact same position, to say nothing of the optics of Chinese generative AI playing a role as well.

I have approached Apple for any comments on the latest SCMP report.

Apple has until the release of iOS 18 in the fall to decide how it best navigates the path between user eagerness for exciting new GenAI tools, and the glaring privacy and security risks from letting AI access user data and from unsecured queries being sent to the cloud for processing, any of which could be sensitive or private.

MORE FROM FORBESWhy DOJ Claims Apple's iMessage Should Change

Apple used the device versus cloud argument in Australia earlier this year, pushing back against a mandate to monitor content. “Tools of mass surveillance,” it warned, “have widespread negative implications for freedom of opinion and expression and, by extension, democracy as a whole… Awareness that the government may compel a provider to watch what people are doing raises the serious risk of chilling legitimate associational, expressive, political freedoms, and economic activity.”

The irony that we now have the potential for cloud-based, Chinese AI on an iPhone is fairly clear. How that AI is policed, how expansive a role it is given, and the privacy protections put in place—even in China, will be critical.

As I have said before, smartphone AI is the next privacy and security battleground. It’s not yet well understood, and user excitement still rules over privacy and common sense. Apple will need to tread carefully as its plans become more clear cut.

Follow me on Twitter or LinkedIn