Apple is preparing to give its voice assistant a major overhaul. Several reports indicate that Siri will evolve from a simple voice assistant to a fully integrated AI chatbot in iOS 27, with capabilities similar to services like ChatGPT or Gemini.
The Cupertino company reportedly decided to accelerate its AI strategy after realizing that Siri's traditional approach had fallen short compared to the competition. With iOS 27, the assistant will evolve into a conversational system capable of holding long dialogues, better understanding context, and performing complex tasks on iPhone, iPad, and Mac.
From limited assistant to conversational chatbot
Until now, Siri has focused on responding to specific commands like setting alarms, sending messages, or playing music , with short and inflexible interactions. Apple's plan for iOS 27 is to break that limitation and offer a continuous conversational experience, where users can chain together questions and requests without starting from scratch each time.
According to sources close to the company and leaks reported by outlets like Bloomberg and MacRumors, the new Siri will have conversational memory : it will be able to remember parts of recent conversations to provide more coherent responses, both by voice and text. However, Apple will impose explicit limits on this memory to protect privacy and prevent indefinite recording of conversations with the assistant.
This evolution represents a significant shift in Apple's approach. Executives like Craig Federighi had insisted that the company didn't want to center the user experience on a "chat" feature , but rather on specific functions that would help in daily life. However, the rise of ChatGPT, Gemini, Claude, and other systems has forced a rethink of that strategy to avoid falling behind.
The new approach will allow Siri to take on a more central role in interacting with the Apple ecosystem , not as an optional add-on, but as a key component for managing tasks, searching for information, and leveraging the AI ​​features of devices.
Project Campos: deep integration into iOS, iPadOS, and macOS
Internally, Apple is developing the new Siri under the codename "Campos ." This project will not be released as a standalone application, but rather as an evolution of the assistant already present in the system. Users will continue to invoke Siri as always: using the usual voice command or by pressing and holding the side or power button, depending on the device.
The biggest difference will be "on the inside." Campos will be natively integrated into iOS 27, iPadOS 27, and macOS 27 , with the ability to work via both voice and text and to interact directly with on-screen content. It will be able to read what the user has open, analyze it, and suggest actions without the need to switch between apps.
This conversational assistant will be present in key applications across the Apple ecosystem. Siri is expected to be able to search and edit photos using only voice commands, compose emails taking into account calendar context , find specific files, summarize documents, generate text drafts, and manage information in apps like Mail, Messages, Music, Photos, and even Xcode for developers.
In the early stages of development, Apple reportedly tested Campos as a separate app, but the company abandoned that approach to maintain the concept of an integrated system assistant . The intention is for the user to perceive Siri as a conversational layer available at any time, and not as an add-on service that needs to be opened separately.
Gemini as the brain: the alliance with Google
To make this leap a reality, Apple won't rely solely on its own technology. Various reports agree that the new Siri will use artificial intelligence models based on Google Gemini , specifically adapted to work within the iOS, iPadOS, and macOS ecosystem.
In the first phase, linked to iOS 26.4, Apple plans to rely on a lighter version of Gemini, known internally as Apple Foundation Models v10 , with approximately 1,2 trillion parameters. This model would power the enhanced version of Siri and some Apple Intelligence features that will run in the company's cloud via Private Cloud Compute.
The real leap forward would come with iOS 27, where Campos would be supported by a more advanced version of Google's models, comparable to Gemini 3 and internally referred to as Apple Foundation Models v11. This model would offer greater reasoning capabilities, better context understanding, and substantially more power to generate and analyze content.
According to leaks, Apple will pay Google around $1.000 billion a year for access to these models , a considerable amount but, in perspective, much less than what Google pays Apple to remain the default search engine in Safari. Even so, the investment in AI will also entail its own infrastructure and maintenance costs that the company will have to manage in the medium and long term.
Apple designs Campos with significant flexibility: the architecture will allow for a change in the AI ​​model in the future if necessary, either to adopt other providers or to adapt to markets with specific restrictions, such as China, where Apple Intelligence is not yet readily available.
Key features of the new Siri chatbot
The future Siri in chatbot form won't be limited to answering simple questions. Reports indicate the assistant will be capable of performing a wide range of advanced tasks by combining natural language, context, and access to user data.
The following are among the planned functions:
- Searching for information on the web, with more elaborate answers and clear summaries instead of simple search results, as suggested by the development of a AI-powered search engine integrated into Siri.
- Content generation: texts, email drafts, message proposals, descriptions, and even help in creating complex documents.
- Image creation and editing using AI, based on voice or text instructions, taking advantage of both generative models and the capabilities of the Photos app.
- Document analysis and summary, including reading files, extracting key ideas, or preparing thematic summaries.
- Help with writing and understanding code, especially integrated into Xcode and development tools, with suggestions, error explanations, and generation of programming snippets.
- Contextual actions on what appears on the screensuch as replying to a long email, completing a form, rearranging notes, or adjusting device settings depending on the task being performed.
- Direct control of system settings and device functions, from changing mobile settings to managing home automation integrated into the Apple ecosystem.
In addition, Siri will be able to access—if the user allows it—personal data stored in native apps to offer more useful answers: locate a specific email, find a photograph just by describing it, review a calendar event related to a conversation, or suggest actions based on information from Messages or Reminders.
All of this will be combined with a dual interaction mode, by voice and by text , so that the assistant can be used both in situations where speaking is comfortable and in environments where writing is preferable, for example at work, on public transport or in meetings.
Privacy, limited memory, and cloud processing
One of the most sensitive issues for Apple is how to reconcile this transformation with its traditional approach to privacy. The company wants the new Siri to be competitive against other chatbots without compromising its image of data protection , something especially relevant for users in Europe, where regulations are stricter.
To achieve this, Apple will employ a hybrid strategy. Some processing will be performed on the device itself whenever possible, while more complex operations will be delegated to external servers controlled by the company , in what Apple calls Private Cloud Computing. Although it draws on Google's models, the infrastructure and data control will remain under criteria established in Cupertino.
Regarding conversational memory, Siri will not indefinitely store all interactions . Instead, it will remember a limited segment of what was said to maintain the immediate context of the conversation. This decision seeks a balance between usefulness and privacy protection, although it could lead to situations where the assistant doesn't remember topics discussed long ago.
Apple is also introducing specific limits on certain sensitive features. For example, Siri's ability to remember personal data will be restricted and configurable, allowing users to decide the extent to which the assistant can access information stored in apps and services.
Two key phases: iOS 26.4 and the definitive leap with iOS 27
The arrival of this new Siri chatbot won't be all at once, but rather in two distinct stages that will be rolled out over the next few months . The first stage will take place with iOS 26.4, an update that will lay the foundation for the future conversational experience.
In iOS 26.4, users will see an improved version of the current assistant, more contextual and personalized , based on the Apple Intelligence features announced in 2024 that had been delayed. This Siri will still use the classic interface, but will be able to better understand requests, analyze what appears on the screen, and search the web more intelligently.
The second phase, planned for iOS 27, will completely transform the assistant into a chatbot integrated across all devices . From that point on, Siri will adopt the new conversational interface, leverage Apple's most advanced Gemini model, and become the central hub of the AI ​​experience on iPhone, iPad, and Mac.
Looking ahead, various reports suggest the official unveiling of these new features will take place at WWDC , Apple's annual developer conference, typically held mid-year. The rollout to the general public would then coincide with the release of new stable versions of iOS, iPadOS, and macOS, traditionally in the fall.
Business model: Is Siri chatbot free or subscription-based?
The fact that Apple is going to pay significant amounts to Google for access to advanced Gemini models has opened a logical debate: will all the functions of the new Siri remain free or will there be a paid tier?
So far, the company hasn't made any official announcement, but the industry offers some clues. Virtually all the major chatbots—ChatGPT, Claude, Copilot, Perplexity, or Gemini itself— operate with a hybrid model that combines free basic use with subscription plans for access to advanced features, more context, server priority, or better models.
In this context, several analysts consider it plausible that Apple will opt for a similar strategy: an improved Siri included for free with iOS 26.4 , accessible to all users, and a more powerful chatbot-type mode, based on higher-capacity models, which could be linked to some kind of subscription when iOS 27 is available.
Google, for example, offers a free version of Gemini integrated into its Android services and devices, while reserving Gemini Advanced for those who pay a monthly fee , with clear improvements in complex tasks, analysis of large documents, and scheduling. Apple could adapt this logic to its ecosystem, leveraging its large user base and prior experience with bundles and services.
One indication of this trend is the upcoming Apple Creative Studio, a suite of professional applications for iPad and Mac that will incorporate advanced AI features exclusively for subscribers. This type of move suggests that, in the medium term, some of the more demanding artificial intelligence capabilities could become tied to paid plans, including Siri.
Europe and the regulatory challenge of AI
In Spain and the rest of Europe, Siri's transformation is taking place within a particularly demanding regulatory environment. The approval of specific regulations on artificial intelligence and the strengthening of data protection mean that Apple has to tread very carefully when deploying features based on external models like Gemini.
In practice, this means the company must ensure that the processing of personal data complies with the GDPR , that users have clear control over what data is sent to the cloud, and that options exist to limit or disable certain features. Privacy settings related to Siri and Apple Intelligence are expected to be particularly detailed in European markets.
Even so, Apple has a certain image advantage over other players in the sector. Its consistent advocacy for privacy and its past experience incorporating regulatory requirements make the rollout of the new Siri in Europe likely, although perhaps with some features reduced or adapted compared to other regions.
European users may encounter specific restrictions on capabilities such as long-term memory or the use of certain external models depending on how EU rules on data transfer and training of AI systems with personal information are interpreted.
Given this landscape, the next major version of Apple's assistant is poised to be one of the most significant changes in the recent history of iOS, iPadOS, and macOS. Siri will evolve from a tool many used only for basic tasks to an integrated conversational chatbot , powered by advanced Gemini models, with the ability to understand context, interact with apps and personal data, and adapt to the privacy requirements of markets like Europe. If Apple manages to balance power, data protection, and a sound business model, the leap that will arrive with iOS 27 could redefine the daily relationship millions of people have with their devices.
