Apple does not need to invent a new place to put a chatbot-style Siri or any future AI assistant. Messages is already the most natural, low-friction surface Apple has for deploying conversational AI at scale, and the groundwork is surprisingly far along already.

Think about how people actually talk to AI today. Most users are not spinning up terminals or developer tools; they are chatting—whether that’s in WhatsApp, Telegram, Discord, Slack, or a dedicated app that still looks like a chat window. Apple has one of the stickiest chat apps on the planet in iMessage, with deep integration across iPhone, iPad, Mac, and Apple Watch. That gives it something OpenAI, Google, and a lot of indie AI apps would kill for: a default, always-pinned place where users already live for hours a day.

The timing is also perfect. With iOS 27, Apple is expected to flip Siri from a static voice assistant into a full-blown AI chatbot, powered by more capable Apple Foundation Models and building on the Apple Intelligence work introduced earlier. Reports describe this new Siri (code-named Campos) as something that can hold real back-and-forth conversations, work across apps, generate content and images, and pull from personal on-device context like messages, emails, and files to actually get things done. Apple is clearly moving into the same territory as ChatGPT and Gemini—just with a different philosophy around privacy and integration.

Right now, the plan that’s been reported is to ship a dedicated Siri “chatbot” interface in iOS 27, replacing the classic Siri UI that has barely changed for over a decade. That’s a big step, but it’s still somewhat siloed: ask a question, get an answer, then leave. Messages, on the other hand, is where conversations live. You already have long-running threads with friends, family, teams, and businesses in there. Dropping conversational AI into that space turns Siri from a thing you invoke occasionally into a contact you just text when you need something.

You can imagine the simplest version of this tomorrow: a “Siri” contact in your Messages list. You text it like you would any other person: “Can you rebook my flight to tomorrow afternoon?” or “Summarize this group chat for me and highlight anything I actually need to do.” The assistant replies inline, can see the context you grant it, and can take actions on your behalf across apps. The biggest UX challenge would be keeping different AI conversations organized, but that’s exactly the kind of thing Apple is good at solving through subtle UI affordances—separate threads, labels, or a picker for different AI “apps” inside Messages, the same way it has contact filters and group naming today.

Where this gets really interesting is when you stop thinking about just “Siri” and start thinking about an entire ecosystem of AI agents living inside Messages. Third-party conversational AI apps already exist on iOS, but right now they mostly feel like standalone islands: you have to remember to open the Claude app for one thing, the ChatGPT app for another, a financial AI app for something else. Apple doesn’t really “own” any of that experience, and developers lose out on the discovery and persistence that come with being in a default system app.

If Apple exposed a proper conversational AI API for iMessage, that dynamic flips. Developers could register AI agents that show up as pseudo-contacts or app-like extensions in the Messages interface. Instead of opening the Claude app to manage your remote coding session, you could just message “Claude Dev” and say, “Restart the RC session from my last project and pull in the new repo,” then watch it work while you keep chatting. A personal finance app could text you, “You’re on track to overspend this month; want me to move $50 into savings?” and you could reply “Yes” without ever thinking about opening the app. Fitness, travel, shopping, productivity—anything that already uses notifications could turn into a conversational agent that actually understands context and can hold a thread over time.

The kernels of this already exist in two places: Messages for Business and the quiet rise of iMessage-based “AI bots” that are running on top of Mac mini server farms. Apple’s Messages for Business (formerly Business Chat) lets you talk to companies over iMessage to get support, schedule appointments, and even pay with Apple Pay, all inside a chat thread. Under the hood, that system supports both human agents and automated flows, and it already integrates with external systems like CRMs, order management, and payment rails. It is, in a sense, Apple’s first large-scale conversational platform—just heavily slanted toward customer service rather than general AI.

Meanwhile, outside Cupertino’s control, indie developers and open-source projects have started wiring up their own AI agents to iMessage using Mac minis as always-on, personal AI servers. Those setups exist precisely because iMessage is such a compelling channel: it’s trusted, end-to-end encrypted, cross-device, and already where your attention is. A Mac mini gives these bots a bridge into that world, letting them respond just like any other contact in your chat list. In other words, the market is telling Apple what people want: AI that talks to you in Messages.

Apple’s job now is to cleanly formalize what is currently a hack. That likely means three layers. The first is the system-level Siri chatbot, fully integrated with Apple Intelligence and Apple’s privacy model, where most of the core capabilities live on-device or in “private cloud” setups that minimize data exposure. This is the assistant that understands your personal context across apps, windows, and documents, and can orchestrate complex workflows without you needing to think about prompts. The second is an extension framework for third-party AI agents that can live alongside Siri inside Messages, with strict sandboxing and permissioning so that each agent only sees what the user explicitly shares. The third is a set of upgraded Business Messages tools so companies can easily swap out clunky decision trees for real LLM-powered assistants while still benefiting from Apple Pay, scheduling, and the existing customer-service infrastructure.

For Apple, the strategic upside of using Messages as the hub is massive. It anchors AI usage directly to the Apple ecosystem instead of ceding attention to standalone chatbot apps or web interfaces. It also reinforces the company’s narrative that AI is not some separate destination but a capability woven into the tools people already use—much like how features such as Live Text or Visual Lookup just quietly appear wherever they make sense. If Siri is just another “participant” in your chats, the AI becomes less of a product, more of a presence. And because Messages is end-to-end encrypted and deeply associated with Apple’s privacy branding, the whole experience gains a level of built-in trust that’s hard for cloud-first competitors to match.

There will be trade-offs. Some power users will prefer the clean separation of a dedicated AI app; some developers will be wary of giving Apple another layer of control over their relationships with users. There are also real UX questions: how do you avoid flooding people’s chat lists with bot threads? How do you keep AI agents from turning into spammy notification machines? How do you design transparent permission prompts that explain what each bot can access and remember, without overwhelming the average user? Those are all solvable problems, but they require careful design and clear rules, not just a new API.

Still, the direction of travel feels obvious. Apple is already on track to ship a much more capable Siri chatbot in iOS 27, backed by upgraded models and deeper OS hooks. Messages for Business has proven that chat as an interface can handle real transactions, scheduling, and support at scale. The developer community has demonstrated that people are willing to go to surprising lengths—up to and including building Mac mini farm backends—just to get AI into iMessage. All that’s missing is the connective tissue.

If Apple chooses to lean into Messages as the home for conversational AI, it won’t just be shipping another chatbot. It will be turning its most human app into the control center for the next decade of computing—where you don’t just talk to people, you talk to software that talks back, understands context, and quietly works in the background on your behalf. For a company that built its brand on making technology feel personal, that might be the most Apple way to do AI yet.