Apple Leadership Clashes Over AI Strategy as Autonomous Agent Race Heats Up

Behind the optimistic facade presented by Apple's new CEO, John Ternus, during his Wall Street debut earnings call, deep concern is quietly brewing within the tech giant's Cupertino headquarters. While Ternus publicly declared that the company has finally delivered the Siri it always envisioned while emphasizing its privacy safeguards, closed-door discussions reveal a starkly different reality. Fully aware that Apple still lags significantly behind its competitors, Ternus recently held a tense meeting with Software Engineering Senior Vice President Craig Federighi, issuing an unambiguous directive to accelerate the company's roadmap and transform the voice assistant into a true autonomous AI agent.

Internal friction and growing frustrations stem from what Bloomberg reports as Federighi's skepticism toward artificial intelligence, driven by concerns over privacy and security implications. Having taken over Apple's AI responsibilities following John Giannandrea's departure last year, Federighi's centralized management style and short-term planning cycles of 12 to 18 months have clashed with the hardware division's five-year strategic plans, triggering internal friction and delays. Mike Rockwell, the visionary behind Apple's mixed-reality headset, is also reportedly weighing a departure next year amid the product's commercial struggles, potentially paving the way for Sebastian Marino or former Google executive Amar Subramanya to step in.

Apple's leadership is increasingly frustrated as the market shifts into the next paradigm. Although the redesigned Siri—released two years behind schedule within iOS 27 and rebuilt using Google's Gemini—can now comprehend on-screen context and retrieve data from emails and messages, it still requires manual user prompts for every action. In the upcoming iOS 27.3 beta, Apple is slated to introduce complex multi-step execution within native apps, followed by third-party app integration this spring for tasks such as restaurant reservations or advanced photo editing. Additionally, the company is attempting to finalize an autonomous tool capable of independently identifying and replacing compromised passwords across the web.

However, compared to rivals in the United States and China, these measures appear overly measured. Advanced systems such as OpenAI's Dots, Meta's Muse, and Elon Musk's Grok already operate as continuous agents managing long-term projects, monitoring background changes, and initiating actions without constant user intervention. Meanwhile, Anthropic and OpenAI offer computer-control tools capable of moving cursors, clicking buttons, and operating software like a human user. In China, manufacturers like Huawei and Xiaomi have already embedded intelligent agents at the operating system level, executing full cross-app automation.

To bridge this widening gap, Apple is fundamentally overhauling the development methodology for its upcoming operating systems, iOS 28 and macOS 28, codenamed Bell and Poppy. Abandoning the rigid cycle of four weeks of coding followed by four weeks of integration, the company has granted engineers six full weeks of open research and development to foster bolder concepts before locking in final features. Apple recognizes that its hardware dominance and custom silicon will not suffice if competitors launch dedicated devices built from the ground up around proactive agents that see, remember, and act independently.

The stakes are particularly high with younger demographics. While Apple's brand loyalty remains robust, Generation Z consumers have grown up in a deeply immersive technological landscape and often bypass Apple's ecosystem intermediaries to adopt cutting-edge innovations. With OpenAI collaborating with legendary former Apple designer Jony Ive, and Google integrating Gemini into hardware ranging from refrigerators to Chromebooks, Apple faces a challenging decade as it strives to maintain relevance through 2036.