Is Apple Struggling with AI?

In the wake of its Worldwide Developers Conference, Apple executives have been on the defensive. They insist that last year’s flashy demo of a supercharged, AI-powered Siri wasn't “demoware,” but rather a feature that just wasn't ready to ship. In an interview with the Wall Street Journal, senior VP Craig Federighi said the company realized it needed a more advanced architecture to meet expectations, pushing the launch to 2026.

This delay comes at a critical time for the tech giant. While executives stress that Apple is playing the long game, aiming to thoughtfully integrate “Apple Intelligence” across its operating systems rather than just building another chatbot, the market is growing impatient. Apple’s stock has underperformed its Big Tech peers, sliding 18% year to date. The delay highlights a larger, more uncomfortable question: Is the company that pioneered the modern smartphone and personal assistant falling behind in the AI race?

What Happened to Apple’s Head Start?

It’s easy to forget that Apple was an early leader. When Siri launched in 2011, it offered a glimpse into a future of voice-controlled computing. But that lead was squandered as competitors like Amazon’s Alexa and Google Assistant quickly surpassed Siri’s capabilities. Internally, Apple’s AI efforts were scattered across different divisions, a problem the company tried to solve in 2018 by hiring Google’s AI chief, John Giannandrea, to unify its strategy.

Yet, nearly six years later, that strategy is being reversed. Here's Mark Gurman of Bloomberg on the situation:

Apple has fallen further behind its tech peers in AI, and promised Siri features keep getting pushed back. Apple Intelligence — launched last year — has had a subpar performance, with many features being late to arrive. It’s clear that putting everything under JG hasn’t worked.

So Apple is starting to break up its AI team and revert to the management structure that existed prior to Giannandrea’s arrival.

This restructuring suggests that even after years of trying, Apple is still searching for the right formula to compete in an AI landscape that has evolved at a breathtaking pace.

Can Apple Just Build Its Way Out?

Apple’s traditional trump card has been its mastery of custom silicon. From the A-series chips in iPhones to the powerful M-series in Macs, Apple’s ability to design hardware and software in tandem has given it a significant performance and efficiency advantage. However, the generative AI race isn't just being run on consumer devices; it’s being fought in massive, power-hungry data centers.

While Apple has focused on efficient on-device processing, its rivals have been pouring staggering sums into AI infrastructure. McKinsey projects that data centers will require nearly $7 trillion in global investment by 2030. OpenAI, in partnership with Oracle and SoftBank, has announced the $500 billion “Stargate” project to build out its own AI factories. This massive server-side compute is essential for training and running the most powerful frontier models, an area where Apple has visibly lagged.

The most telling sign of this gap came in March 2025, when reports surfaced that Apple was buying Nvidia GPUs for its own AI data centers for the first time in nearly two decades — a tacit admission that its own hardware isn't enough to power its AI ambitions.

Is It a Software and Talent Problem?

Hardware is only one part of the equation. Nvidia’s nearly unassailable position in AI is built on CUDA, a proprietary software platform that has become the industry standard for GPU programming. For almost two decades, an entire generation of AI researchers and developers has been trained on CUDA, creating a deep and sticky ecosystem that competitors like AMD have found nearly impossible to replicate.

Apple has its own powerful frameworks like Metal and Core ML, but they are confined to the Apple ecosystem and have less traction in the high-performance computing (HPC) world where cutting-edge AI is born. Recognizing this, Apple announced at its developer conference that it would open its foundational AI models to third-party developers, a clear attempt to leverage its massive App Store moat to build a new, AI-specific one.

The talent war further complicates things. Top AI researchers are being offered multi-million dollar compensation packages, with one top mind reportedly commanding $40 million a year. This intense competition for a small pool of elite talent makes it difficult for any company, even Apple, to catch up overnight.

How Are Rivals Seizing the Opportunity?

While Apple recalibrates, its competitors are moving aggressively. OpenAI, not content with its software lead, made a stunning move into hardware by acquiring io, the startup founded by legendary Apple designer Jony Ive, in a $6.5 billion deal. Their stated goal is to create a new “family of devices” and ship 100 million “AI companions,” with the first product slated for 2026. Ive, the design mind behind the iPhone, is now working with Apple’s biggest AI rival to create a product that could one day threaten it.

Meanwhile, Google is leveraging a massive, under-the-radar advantage: its custom Tensor Processing Units (TPUs). By designing its own AI chips, Google avoids paying the hefty “Nvidia tax” — estimated at 80%+ gross margins on data center GPUs — allowing it to run its AI workloads at what analysts suggest is just 20% of the cost incurred by its rivals. This economic advantage allows Google to offer its powerful Gemini 2.5 Pro model, with its massive one-million-token context window, at aggressively low prices.

Even Meta is finding success. While its Llama models push the open-source frontier, its Ray-Ban smart glasses have become a surprise hit, selling over 2 million pairs and demonstrating a viable path for AI-infused consumer hardware.

What's Next for Apple?

Apple's stated strategy is not to win the chatbot wars but to weave intelligence seamlessly into the user experience. The partnership to integrate ChatGPT into its ecosystem is a crucial stop-gap to address its current model deficiencies. However, the company is fighting a multi-front war. The very foundation of its services revenue is under threat from antitrust lawsuits targeting its App Store fees and its lucrative default-search deal with Google.

Apple’s own services chief, Eddy Cue, recently admitted that AI could create a technology shift that makes the iPhone obsolete within a decade. With its most iconic designer now working for the competition and its core business models under regulatory fire, Apple is in the unfamiliar position of playing defense. The company has a famously loyal user base and a powerful platform, but in the fast-moving AI era, that may no longer be enough to guarantee its place at the top.

Reference Shelf:

Apple Executives Defend Apple Intelligence, Siri and AI Strategy (WSJ)

Apple executives say new AI-powered Siri wasn’t ‘demoware’ — it just wasn’t ready to ship (TechCrunch)

Apple's AI Reboot: Wooing Developers in a High-Stakes Race (ARPU)