Skip to content
IFAI

Industry

Apple’s AI playbook for iOS 18 starts coming into focus

Published Reach 37

Apple is reportedly preparing a suite of AI features for iOS 18 powered by its Ajax LLM, including text summarization in Messages and Safari. Simple tasks will run on-device with strict privacy warnings, while complex queries could tap server-side power from Google or OpenAI. Details are expected at WWDC on June 10th.

Apple’s AI playbook for iOS 18 starts coming into focus
Image: theverge.com

Apple CEO Tim Cook has been promising big AI moves for months, and now we are finally getting a clearer look at what that means for your next iPhone.

According to a new report from AppleInsider, iOS 18 is set to give Siri a massive functional upgrade. The voice assistant will reportedly gain the ability to analyze and summarize text conversations directly inside the Messages app.

On-device brains meet server muscle

The text summarization feature will be powered by Ajax, Apple’s own large language model.

It matches up with earlier reports that Safari in iOS 18 is getting an Intelligent Search tool designed to summarize web pages. Instead of reading a massive article, users will be able to get a quick recap in seconds.

Apple has published several research papers pointing toward on-device processing, and the reporting confirms that Ajax will handle simple requests entirely on the iPhone itself.

However, local mobile hardware has its limits. When a prompt gets too heavy, Ajax will reportedly route the query to external servers. That explains why Apple has been exploring partnership deals with both Google and OpenAI to handle more demanding generative tasks.

Privacy will remain front and center. Apple will reportedly display clear privacy warnings before Ajax accesses sensitive personal data from Safari or Messages.

Cook reiterated on Thursday’s earnings call that Apple feels very bullish about AI and has an unwavering focus on privacy. While Apple has an event next Tuesday for new iPads, the full AI reveal will almost certainly wait until WWDC on June 10th.

Why it matters

Apple is not trying to invent another standalone chatbot. Instead, it is quietly integrating AI into the basic communication tools you already open fifty times a day.

For everyday users, getting a fast, clean recap of a chaotic group chat or a long browser tab right on the device is far more practical than having a philosophical conversation with an AI. If it runs locally, it means zero latency and no fear of personal data leaking onto public servers.

We think leaning on a hybrid setup is a clever pragmatic move. Handling everyday text tasks locally keeps Apple’s compute bills low, while outsourcing heavy reasoning to players like Google or OpenAI buys Apple time to build out its own cloud models.

However, the privacy prompt could become a friction point. If users are constantly interrupted with warnings every time a task gets too complex, it might break the seamless illusion Apple products usually offer. Either way, June 10th is shaping up to be the most consequential WWDC in years.

Sources (1)

  1. [1]
    theverge.com(opens in a new tab)Originaltheverge.com

Topic trackerApple's Generative AI Strategy

Every story on this topic, newest first.

  1. Apple walks away from OpenAI funding round
  2. Apple in talks to invest in OpenAI
  3. Microsoft and Apple walk away from OpenAI's board
  4. Apple lands a seat at the OpenAI table
  5. Apple could add Google Gemini to iOS this fall

Best of AI, daily