Apple-Google Siri tie-up signals AI sector maturity, dominance of a few

Apple's move to use Google's Gemini to power Siri highlights how rising costs and complexity in AI are pushing even the biggest tech firms towards collaboration and reliance on a few dominant players

iPhone-maker Apple on Monday said it would use Google’s large language model (LLM), Gemini, to power Siri. The $1 billion-a-year, multi-year deal will also see the next generation of Apple foundation models based on Gemini models and cloud technology. 

In a joint statement, both companies said that Google’s AI models will help power future Apple Intelligence features, including a more personalised Siri, which is likely to be available on the new iPhone and other Apple devices slated for release in April or May this year.

 

“After careful evaluation, Apple determined that Google’s AI technology provides the most capable foundation for Apple Foundation Models and is excited about the innovative new experiences it will unlock for Apple users. Apple Intelligence will continue to run on Apple devices and Private Cloud Compute, while maintaining Apple’s industry-leading privacy standards,” the two companies said.

Apple’s AI strategy under scrutiny

Though this move is being seen as Apple raising the white flag in the AI race, some, such as Elon Musk, also see it as an “unreasonable concentration of power for Google” since the company also offers Android and Chrome.

 

Some experts believe this move underscores the view that Apple’s internal AI development has fallen behind competitors like Amazon, Meta, and Microsoft, despite early promises at WWDC 2024.

 

“Apple struggled to deliver on its AI roadmap, delaying the Siri upgrade and ultimately concluded that its own models weren’t reliable enough. This multi-year AI partnership with Google is a strategic pivot, allowing Apple to rapidly close its AI gap while protecting its privacy-focused brand and financial margins,” said Salman Waris, founder and managing partner of law firm TechLegis.

Why Apple’s approach to AI differs from rivals

Other experts, however, say that Apple’s use of Gemini to upgrade Siri’s capabilities should not be interpreted simply as a shortfall in Apple’s technical competence, but rather as a reflection of how the company has historically approached AI.

 

“Unlike peers that have prioritised rapid deployment of large-scale generative models, Apple has emphasised on-device processing, privacy preservation, energy efficiency, and deep integration across hardware and software ecosystems. These design constraints, while aligned with Apple’s brand and regulatory posture, inevitably slow the pace at which frontier-scale language models can be developed and deployed internally,” said Jameela Sahiba, associate director, The Dialogue, a tech and public policy think tank.

Rising costs push firms towards AI partnerships

Some other experts are also of the view that the partnership with Google is a more practical and pragmatic approach to how companies will think about AI going ahead.

 

“The costs of training frontier AI models may have come down from their peaks, but if companies are to retain their edge in computing, the need to spend billions of dollars, both in research and then in product development, will not go away. It is then natural that not all companies will invest so much and instead rely on others’ progress to power their own products,” an industry expert said, asking not to be named.

AI industry enters a phase of consolidation

Apple’s decision also illustrates a maturation phase in the AI industry, where the complexity, cost, and scale of frontier model development are beginning to exceed the efficiency of fully siloed innovation, even for the most resource-rich firms, Sahiba said.

 

“Training and maintaining state-of-the-art foundation models now demands massive compute infrastructure, highly specialised research teams, and continuous model iteration, creating structural incentives for collaboration, licensing, and modular integration. This does not imply a convergence towards homogeneity or a loss of strategic independence,” she said.

 

Overall, the broader implication for the AI industry is that it will move towards integrated ecosystems, where core AI capabilities are outsourced to a few dominant providers, Waris said.

 

“More companies will increasingly collaborate rather than build everything in-house, especially in AI. The trend towards competition where competitors collaborate in some areas while competing in others is accelerating due to the complexity, cost, and scale of modern AI systems,” he said.