Enhanced Privacy and Speed: iOS 18 Introduces On-Device AI Features with LLM Technology

Apple is heavily investing in artificial intelligence technologies, with plans to announce its AI strategy in June at WWDC. The upcoming iOS 18 and other new OS versions will showcase these advancements.

Mark Gurman’s Power On newsletter hints at the new iPhone AI features being powered by an offline, on-device large language model developed by Apple. This approach is expected to highlight privacy and speed benefits.

Previous discoveries in iOS 17.4 hinted at an on-device model named “Ajax,” with Apple also working on server-hosted versions of Ajax. While on-device LLMs may not match the power of models on huge server farms, Apple’s integration of software and silicon chips can optimize performance.

On-device LLMs offer faster responses and offline capabilities, advantageous in areas with limited connectivity. They may not have the same breadth of knowledge as cloud-based models but can excel in specific tasks like auto-replies and Siri requests.

Apple’s focus on privacy aligns well with on-device AI models, ensuring data stays local. These models can also handle generative tasks like document or image creation effectively.

Apple’s WWDC keynote on June 10 will reveal its full AI strategy, showcasing new software features for various devices. Stay tuned for updates on how Apple plans to leverage AI advancements.