Apple’s AI Breakdown: Why Siri Failed and Google Is Now Saving It

How Apple Fell Behind in AI—and Why It’s Paying Google $1 Billion to Catch Up

Back in December 2023, Apple executives Johnny Shroudy and John Turnus confidently dismissed concerns that the company was lagging in artificial intelligence.

Fast forward to today — and that confidence feels misplaced.

Siri, Apple’s once-groundbreaking voice assistant, now trails far behind its competitors. It’s not just outdated — many would say it’s broken. In a meeting held in March 2025, Robbie Walker, a senior director at Apple, reportedly described the state of “Apple Intelligence” as “ugly and embarrassing.” He even called the company’s decision to promote unfinished AI technology “an absolute disaster.”

The AI features that were heavily advertised with the iPhone 16 — Apple’s supposed leap into the AI era — never actually arrived. And now, in an unexpected twist, Bloomberg reports that Apple is turning to an unlikely ally: Google.

A Billion-Dollar Surrender

According to Bloomberg, Apple plans to pay Google $1 billion per year to use a custom version of its Gemini AI model to power Siri. The special Gemini system will run on Apple’s private cloud to maintain privacy and security — but the move signals something Apple rarely admits: defeat.

It raises two critical questions.
First, how much does AI really matter in a smartphone? Do users even care?
And second, if a company can simply lease an AI model, why spend billions building one from scratch?

These are questions that challenge not just Apple’s strategy, but the entire AI arms race.

Do People Even Want AI on Their Phones?

When people shop for a new phone, they usually think about price, camera quality, and battery life — not artificial intelligence.

According to a CNET report, only 11% of U.S. smartphone users upgraded their devices because of AI features in 2025 — a 7% drop from the year before. Nearly 30% said they don’t find mobile AI useful at all and don’t want more of it.

Samsung learned this lesson too. Despite marketing its Galaxy AI as the next big thing, most customers didn’t upgrade for it. On earnings calls, Samsung’s focus quickly shifted back to “market weakness” and “economic uncertainty.”
The message is clear: AI in phones is a nice bonus, not a dealbreaker — at least, not yet.

Siri’s Long Fall From Grace

You’d think a company with Apple’s resources — endless cash and armies of engineers — could fix Siri. But Siri’s decline has been public and painful.

Since the mid-2010s, Siri has lagged behind Google’s Assistant, and the gap has only widened. Try a simple test today: ask both Siri and Google Gemini to compare two phones. Gemini delivers a clear, detailed breakdown. Siri? It gives you a handful of Google links.

Even for basic commands — like “play the Friends theme song” — Siri fumbles, while Gemini nails it instantly. For Apple, Siri has turned into a long-running embarrassment.

In 2024, Apple tried to reset expectations with the grand unveiling of “Apple Intelligence.” But instead of delivering smart, contextual, in-app actions, users got the same old Siri — one that now leans on ChatGPT for writing, image generation, and simple on-screen analysis. Essentially, Siri is doing less than ever.

And when delays hit again, Apple pushed the new Siri launch to mid-2026.

Chaos Inside Apple’s AI Division

Behind the scenes, things were worse than anyone realized.
By mid-2025, Apple’s Head of Foundational AI, Ruming Pang, had left for Meta. His departure triggered a wave of resignations among top engineers. Internal reports described the AI division as split into rival camps — one working on Siri, the other on core AI models — constantly clashing and blaming each other.

The Siri team reportedly took two years just to remove “Hey Siri” from its activation phrase. When ChatGPT launched in late 2022, Apple engineers scrambled to catch up — and never quite did.

Apple’s obsession with perfection backfired: generative AI was simply too unpredictable for Apple’s tightly controlled ecosystem. When Apple Intelligence was announced in 2024, it wasn’t ready. Some insiders even called the announcement “false advertising,” and internal finger-pointing soon followed.

When talks with Anthropic for a possible partnership collapsed (reportedly over money), Apple was left with few options. That’s when Google entered the picture.

Why Apple Chose Google’s Gemini

Apple’s deal with Google isn’t just about survival — it’s about buying time.

The Gemini model that Apple will use is said to have 1.2 trillion parameters, compared to Apple’s modest 150 billion. In simple terms, it’s far more capable. And at $1 billion a year, it’s not a terrible deal for Apple — especially since Google already pays Apple around $18 billion annually to remain the default search engine on iPhones.

The plan is for Gemini to handle Siri’s summary generation and multi-step planning, running entirely on Apple’s private servers to maintain its “privacy-first” image. Apple will likely avoid mentioning Gemini in any marketing materials, keeping the partnership low-profile.

The Bigger Picture: AI, Cost, and Control

Apple has always been a “fast follower” — arriving late to a trend, but executing it better. The company might simply believe that AI isn’t mature enough yet for Apple-level polish. As one executive put it,

“It just doesn’t work reliably enough to be an Apple product.”

But here’s the dilemma: investors and shareholders expect Apple to compete in AI. So, even if Apple isn’t ready to fully commit, it has to show progress — and Google’s Gemini provides a convenient bridge until Apple’s own model catches up.

This brings us to a broader insight about the AI industry.
Perhaps not every company needs to spend trillions building its own AI model. Think of the ecosystem like a smartphone:

• Hardware (AI data centers) = Infrastructure
• Operating system (AI models) = Foundation
• Apps (AI-powered features) = What users actually interact with

Maybe the real opportunity lies not in building the model, but in creating great applications that use it — just like Apple uses Gemini to power Siri.

So, What’s Next for Apple?

In the short term, Apple will be fine. iPhone 17 sales in China are up nearly 29% year-over-year, and MacBooks continue to gain market share thanks to their M-series chips and impressive battery life. Even with this AI setback, Apple’s business remains rock solid.

Still, the partnership with Google marks a rare moment of vulnerability for the world’s most valuable tech company. It’s a reminder that even Apple — the master of control — can’t control everything, especially when it comes to the fast-moving world of AI.

So, what do you think?
Is AI in smartphones overrated?
Should companies stop chasing the AI hype and focus on real-world features instead?

---

Follow StoryAntra for more stories like this — where tech meets truth and every detail unfolds with depth.