Apple recruits engineers to develop agentic AI assistants
Apple is seeking engineers to build AI agents capable of web browsing and app interaction. Discover what this means for the future of agentic AI.
Source: The Verge — AIApple is actively recruiting engineering talent to develop its own proprietary AI agents. This strategic move aims to transition the company's AI capabilities from simple conversational responses to autonomous agents capable of executing complex, multi-step workflows.
What is Apple building with these new hires?
According to a job listing recently identified by Bloomberg and reported by The Verge, Apple is searching for engineers to lead the creation of "agentic AI assistants." Unlike standard chatbots that primarily generate text, these agents are intended to function as active participants in a user's digital environment.
The job description specifies that these agents will be designed to plan and execute tasks across several different interfaces, including web browsers, command lines, APIs, and various applications. A key component of this vision involves the ability for these primary agents to coordinate specialized "sub-agents" to complete intricate processes. This suggests a hierarchical approach to task management where a central intelligence delegates specific technical or creative sub-tasks to more specialized models.
How will these AI agents function in practice?
In practical terms, an agentic assistant from Apple would move beyond answering questions to performing actions. The job listing notes that users currently perform valuable work by moving between tools such as editors, terminals, and browsers; Apple's goal is to build AI that can replicate this fluidity.
Specific capabilities mentioned in the listing include:
- Web Browsing: Navigating the internet to find information or complete transactions.
- App Interoperability: Working across different software applications to move data or trigger actions.
- Code Execution: Writing and potentially running code to solve problems.
- Service Integration: Calling external services to fulfill user requests.
This development follows Apple's recent updates to Siri, which began rolling out enhanced AI capabilities in September. While current iterations allow users to ask questions about on-screen content, the new direction points toward deeper integration, such as the ability to organize Safari tabs or perform photo editing tasks like cropping and rotating directly through voice or text commands.
Who are Apple's main competitors in the agent space?
By pursuing agentic AI, Apple is entering a highly competitive landscape populated by several major technology players. The Verge notes that Apple will be competing with established and emerging agentic tools, including:
- Meta's Muse
- OpenAI's Dots
- Anthropic's Duet
- Google's Tasks
Several of these competitors have already made strides in "computer use" capabilities. For instance, Mac applications for ChatGPT and Perplexity can already take certain levels of control over a computer, and Google has begun implementing similar functionality within its Gemini ecosystem. Apple's advantage likely lies in its deep integration with its own hardware and operating systems, potentially allowing for more seamless, low-latency agentic actions across the iOS and macOS ecosystems.
How does this impact the developer ecosystem?
Apple appears to be preparing the ground for third-party developers to participate in this agentic shift. With the upcoming iOS 27.3 update, the company is reportedly introducing features that allow developers to integrate Siri AI into "all apps."
Currently being rolled out via a Developer Preview, this functionality is intended to eventually allow Siri to perform tasks across a much broader range of third-party applications. This indicates that Apple's vision for AI agents is not limited to its own first-party software, but rather a platform-wide intelligence that can act as an orchestrator for the entire mobile and desktop experience.
How to explore agentic AI today with Grafy
While Apple is still in the hiring and development phase for its agentic assistant, the concept of an AI agent that can navigate the web and perform tasks is already a core part of the Grafy ecosystem.
Users interested in seeing agentic behavior in action can use Operator, which functions as a browsing agent designed to perform web-based tasks. For those interested in the technical side of how agents interact with code, Grafy Code provides an AI coding studio environment. If you are looking to experiment with the underlying models that power these types of assistants, Grafy Chat allows you to interact with various models side-by-side to compare their reasoning and task-handling capabilities.
Frequently asked questions
What kind of tasks can Apple's new AI agents perform?
Apple's AI agents are designed to carry out multi-step tasks by working across different platforms. According to the job listing, they will be able to browse the web, work across various apps, write code, and interact with APIs and command lines. They may also coordinate specialized sub-agents to handle specific parts of a complex request.
Which companies is Apple competing against in the AI agent market?
Apple is entering a market with several prominent competitors. The Verge reports that Apple will be competing with products like Meta's Muse, OpenAI's Dots, Anthropic's Duet, and Google's Tasks. Some of these competitors already offer capabilities that allow AI to interact directly with computer interfaces.
How will developers interact with Apple's agentic AI?
Through the upcoming iOS 27.3 update, Apple plans to allow developers to integrate Siri AI into their own applications. This feature, which is being introduced in a Developer Preview, is intended to eventually enable Siri to perform actions and tasks across a wide variety of third-party apps rather than just Apple's own software.
Source: The Verge — AI — Apple is hiring engineers to build its own AI agent


