Apple introduced its seemingly revolutionary “Apple Intelligence”, last year, at WWDC 2024. The company promised a brand new Siri. A revamped AI assistant that was aware of contexts, capable of long-term memory storage, and highly personalized. It was meant to be Apple's entrance into the AI sector, which has been primarily dominated by companies like OpenAI and Google. What made this personal assistant so useful was Apple's access to user data. By tracking users' activity, on their phones across their apps, this AI assistant would reach a level of personalization and efficiency that was unseen, yet, in other products.
However, what followed was an inability for Apple to follow through. Most of the features revealed during this conference remain unreleased. The deeper issue isn’t that Apple Intelligence is behind schedule. It’s that the AI landscape no longer rewards the things Apple optimizes for, like secrecy, vertical control, and waiting until a product is perfect. The game changed, while Apple remains bound to its ways.
The Privacy Tradeoff
Apple’s on-device AI strategy is built around privacy. Its models run locally on Apple Silicon chips, not cloud servers. It’s a bet that people care more about keeping their data safe than about how clever their AI seems. It’s a noble theory, but it’s not how most people currently use technology.
What Apple won’t say out loud is this: on-device AI means worse AI, at least for now. Smaller models, fewer capabilities, slower updates are all tenants of on-device AI. And for consumers, these new digital tools are for trivial tasks, meaning data privacy is not a top concern. They’re asking why it still can’t schedule a meeting, summarize a PDF, or hold a conversation that lasts more than one interaction.
It’s not that privacy doesn’t matter. It’s that Apple keeps designing for a user who doesn’t exist, yet: one who prioritizes abstract safety over actual functionality. The result is a Siri that feels five years behind its competitors.
So, while competitors are continuing to launch faster, better AI models, Apple is yet to even fully enter the space. For the majority of AI developing corporations, the playbook was a reliance on user interactions. They were able to improve their AI systems by being used, broken, and patched. However, Apple doesn't work that way. It waits, and follows the traditional pathway of always releasing a polished, finished product. So, it's not that Apple is incapable of building an artificial intelligence model in the competitive landscape. Rather, it's Apple's perfectionism and unwillingness to release anything that it can't fully control that has been holding it back.
What’s Next?
As the initial hype that has surrounded AI continues to settle down, the things users look for has started to shift. The public is no longer captivated by innovations and shiny new upgrades different AI systems are able to provide; they’re beginning to ask which companies are to be trusted with their data, as their reliance on AI assistants grows. 86% of Americans say data privacy is a growing concern, with only 21% trusting the current leading companies to use data responsibly. So, while the early months of the AI race was rewarding experimentation and the speed in which updates were being delivered, these systems are all becoming more normalized.
In that transition, Apple may be able to quietly reassert its relevance. Its AI strategy, which has been pioneered by its deliberate and privacy-first approach, reflects the type of company that users would be willing to switch over to, even if they've grown accustomed to another product. Apple does not need to dominate the conversations about AI innovation. It needs to offer a product that aligns with the values that the public already attributes to the legacy brand, like intuitiveness and safety.
The era of Apple isn't over. It's evolving. In a world that is considering the question of who we trust with our data, Apple's most competitive advantage might be its intentionality.