Apple's rebuilt Siri finally arrives next month, powered partly by Google's Gemini models, with a critical advantage over ChatGPT and Claude: it can search and reason across your personal data like messages, emails, photos, and calendar entries. After two years of broken promises and false starts, the company is ready to deploy what it calls Siri AI, an entirely new version built on Apple Intelligence that transforms the assistant from a tool good at timers and alarms into a conversational system that understands context and holds multi-turn conversations.

What Makes Apple's New Siri Different From ChatGPT and Gemini?

The fundamental difference between Siri AI and competitors like ChatGPT, Google Gemini, or Anthropic's Claude comes down to one thing: access to your personal information. ChatGPT does not know what is in your inbox. Siri AI, with the user's permission, does. This capability, which Apple calls "personal context understanding," lets the assistant handle the mundane but frustrating tasks that currently require manual scrolling through old messages and emails. Finding a restaurant recommendation buried in a months-old text message, pulling a confirmation number from a hotel email, or retrieving photos from a specific trip all become instant voice commands.

Apple has been candid about where the underlying model capability came from. The company's developer announcement states that the next generation of Apple Foundation Models was custom-built in collaboration with Google and its Gemini models. Apple is buying model capability while maintaining control over the device, the app interfaces, the personal index, and the privacy layer. This partnership represents a notable shift for Apple, which has historically developed its own AI systems in-house.

How Does Siri AI Actually Work on Your iPhone?

• On-Device Processing: Siri AI runs on smaller foundation models stored directly on your iPhone, plus larger models on Apple servers through a system called Private Cloud Compute, which Apple says does not store or make personal data accessible to the company.
• Personal Context Search: The assistant can search and reason across your messages, emails, photos, notes, and calendar with your permission, with the index residing on the device rather than on a server.
• App Integration: Third-party apps can participate through App Intents, a framework that lets developers declare what their app contains and what it can do, making content discoverable and operations invokable by Siri.
• On-Screen Awareness: Siri can understand and act on whatever is currently displayed on your screen, collapsing familiar sequences like finding an app, opening it, navigating menus, and performing an action into a single spoken sentence.
• Multi-Device Sync: A dedicated Siri app syncs conversation history through iCloud, so a question asked on a Mac can be picked up on an iPhone, and users can type or speak their requests.

The experience collapses what used to be a multi-step process into a single command. Instead of opening your banking app, navigating to a specific transaction, and manually entering details, you can simply ask Siri to perform the action. However, until banking apps, travel tools, and productivity software adopt App Intents properly, the experience will likely feel incomplete.

Which iPhones Will Get Siri AI, and When?

Siri AI does not run on all iPhones. Apple lists it for iPhone 16 models or later, plus iPhone 15 Pro and iPhone 15 Pro Max. The most advanced on-device model, which drives expressive voices and improved dictation, requires 12 gigabytes of unified memory, a threshold that currently means the iPhone Air, iPhone 17 Pro, and iPhone 17 Pro Max. Supply chain analysts expect the upcoming iPhone 18 Pro and iPhone 18 Pro Max to include 12 gigabytes of memory, meeting this requirement.

Apple is expected to hold its iPhone event in the first half of September, with the iPhone 18 Pro, iPhone 18 Pro Max, and a first foldable model, possibly called the iPhone Ultra, all anticipated to launch. South Africa is on the initial English language list for Siri AI, alongside Australia, Canada, India, Ireland, New Zealand, Singapore, and the UK. South African users also avoid the Digital Markets Act problem that may keep Siri AI off iPhones and iPads in the European Union at launch.

How Does Google Gemini Fit Into Apple's AI Strategy?

The collaboration between Apple and Google represents a pragmatic approach to AI development. Rather than building all model capability in-house, Apple is leveraging Google's Gemini expertise while maintaining control over the user experience, privacy protections, and device-level integration. The same logic extends to the Foundation Models framework, which now lets developers call Apple's models or, through a new language model protocol, models from other providers including Claude and Gemini. A separate framework called Core AI lets developers run their own models locally on Apple silicon.

This openness contrasts with Apple's historical approach and suggests the company recognizes that no single vendor can dominate every aspect of AI development. By partnering with Google on foundational model capability, Apple can focus its engineering resources on the differentiating factors: personal context understanding, on-device processing, privacy architecture, and seamless integration with the iPhone ecosystem.

The measure of Siri AI will not be whether it out-talks ChatGPT or matches Gemini's general knowledge capabilities. Instead, it will be whether an assistant that already holds your messages, your photos, and whatever is on your screen proves more useful than one that holds none of it. No app downloaded from the App Store starts from that position. Apple is coming from behind after years of delays, but the company has a unique advantage: the phone in your hand and everything on it.