According to Bloomberg reporter Mark Gurman in the latest edition of the “Power On” newsletter, Apple’s new chief executive John Ternus has held closed-door meetings with software engineering chief Craig Federighi in recent weeks, explicitly demanding that he accelerate the pace of the artificial intelligence business and deliver a more aggressive plan than “completing the necessary Siri feature upgrades,” with the goal of transforming Siri into an AI agent capable of completing tasks on users’ behalf autonomously.
The internal pressure stands in sharp contrast to Ternus’s public remarks. On last week’s earnings call, the new chief executive offered high praise for Apple’s AI business, saying the new version of Siri had finally brought to life what the company had envisioned for years, that the AI business had opened a new chapter, and that “we’re very pleased with the reviews and feedback we’ve received, and we’re very excited to see more and more users able to experience it.” But according to the report, internally he judges that Apple remains clearly behind its rivals in the AI race, and the optimistic tone does not reflect the real pressure.
Federighi took over the relevant business last year after Apple’s former AI chief John Giannandrea failed to deliver on Siri promises, and delivered a working new version of the assistant in a relatively short period, and is seen as someone capable of cleaning up situations other executives struggle to handle. Yet he is still regarded by outsiders as an AI skeptic, with concerns about privacy, security, and the social and employment impact. The report says Ternus is dissatisfied that he has not embraced AI as aggressively as competitors have.
The industry environment no longer allows Apple to stand on the sidelines. The AI wave led by OpenAI, Meta, Google, Anthropic and others is reshaping device entry points, and the room Apple once had to opt out has essentially disappeared.
Beyond strategic pace, the management style of Apple’s software division has also been brought into focus. Federighi is accustomed to deep involvement in meetings and major decisions, and relies on his grasp of technical details and debating ability to rebut proposals he disagrees with, leading some employees to feel that dissenting views are hard to get support and that the software engineering and design teams have limited influence over the final direction of products. On planning cycles, he typically sets roadmaps only 12 to 18 months out, while the hardware division is accustomed to laying plans years in advance; the two rhythms sometimes fail to align, which both slows development and may make first-generation products more conservative in features and design.
A key task facing Ternus is precisely to align the development rhythms of hardware and AI software. Apple remains an industry leader in hardware design and engineering, but the pace of software iteration does not always keep up.
Uncertainty at the personnel level is also mounting. The report says Mike Rockwell, the creator of the Vision Pro headset who also oversees Siri AI development, is still seriously considering stepping down, and may give up the vice president of Siri engineering role next year or even leave Apple. Given his role in finally bringing the new Siri to market after two years of delays, such a change, if it materializes, would be a major loss.
To speed things up, Apple is also experimenting with new development processes that give engineers 50% more time for coding and exploration and reduce investment in integration, an adjustment directly tied to improvements in coding-agent capabilities.
From a market perspective, Apple’s AI narrative is at a critical point of shifting from “catching up on the basics” to “chasing the leaders.” Investors had grown used to viewing Apple as an AI laggard, and management’s optimistic language on the earnings call had propped up valuation expectations; if the internal push is indeed intensifying as the report describes, Siri’s evolution toward an agent will touch multiple business lines including voice assistants, app distribution and on-device computing power. But adjustments in management style, the磨合 of planning cycles and the risk of losing core personnel all mean variables remain at the execution level. For the supply chain and developer ecosystem, whether Apple can convert its hardware advantage into an AI experience advantage will determine its say in the next upgrade cycle.