iOS 18 AI Will Use On-Device Processing To Preserve Privacy

Apple’s AI efforts in iOS 18 will focus on privacy by performing the required processing directly on the iPhone, not on cloud servers, according to a new report from Bloomberg’s Mark Gurman. 

Over the past few months, we’ve seen several reports about Apple’s upcoming Artificial Intelligence efforts. The first fruits of its labor are expected to debut with the release of iOS 18 and macOS 15 later this year with Apple widely expected to release AI-enhanced features in the new operating systems.

Bloomberg‘s Mark Gurman wrote in the latest edition of his weekly PowerOn newsletter on Sunday that the initial group of AI features will be performed on the iPhone or Mac without requiring an internet connection. Gurman says the initial AI features in iOS 18 “will work entirely on device,” meaning the new features will be performed with no need for an internet connection or cloud-based processing.

Meanwhile, AppleInsider reports that “individuals familiar with the matter” tell them that the report’s claims are accurate. Apple has been reported to be working on its own in-house “Ajax” large language model, or LLM. While more advanced AI features will eventually require a connection to the Internet, basic text analysis and response generation should be available even if the device is offline.

We’re expecting to see Artificial Intelligence features added to several Apple apps, including Siri, Spotlight Search, Safari, and Messages. The publication says Apple has been testing the generation of text-based responses for some time now, so that feature will most likely be one of the first features to debut.

While Apple’s initial AI features may initially be processed on-device, future AI-related features will still require an internet connection to allow cloud processing. Numerous reports indicate that Apple has been negotiating with Google and OpenAI to use their AI technology for iOS 18 instead of developing its own online large language model.

AI tools have become increasingly a part of users’ lives over the past year, sparking Apple’s interest in artificial intelligence. These tools have also become significantly more accessible to consumers while raising questions about users’ privacy due to the on-cloud processing of many AI features.

AI development has become a priority at Apple as it seeks to compete with other companies’ chatbot services and as AI functions in apps have grown rapidly over the last few months.

Users have also expressed concerns about AI “hallucinations,” as models seemingly invent information off the top of their virtual heads. Since AI models feed on content generated by other AI models, the problem continues to grow. Some reports indicate AI models are seemingly growing dumber due to this issue.

Apple’s “Ajax” LLM is intended to compete with other AI tools currently in use. Apple’s efforts to eliminate the cloud-based processing requirement while also improving the text generation quality could give Ajax an advantage over the AI tools and services offered by competitors.

We’ll likely see Apple unveil its fledgling AI efforts and outline its future AI plans during this year’s Worldwide Developers Conference (WWDC), which kicks off on June 10.

[The information provided in this article has NOT been confirmed by Apple and may be speculation. Provided details may not be factual. Take all rumors, tech or otherwise, with a grain of salt.]