Apple Intelligence took center stage at WWDC 2025 as Apple opened its 3-billion parameter on-device AI models to developers. This bold move promises a new era of fast, private, offline AI apps for iPhones, iPads, and Macs. However, as Apple empowers developers, the persistent challenges with Siri and the limited scale of its AI raise critical questions about the company’s position in the AI race

On June 9, 2025, Apple launched its highly anticipated Worldwide Developers Conference (WWDC) in Cupertino, unveiling a transformative update to its Apple Intelligence platform. The centerpiece of this year’s conference is the company’s bold move to open its AI technology to developers, enabling them to harness on-device AI models with approximately 3 billion parameters. This strategic shift promises to reshape the app ecosystem for devices like iPhones, iPads, and Macs, but it also arrives under a cloud of scrutiny regarding Apple’s broader AI ambitions, particularly with Siri’s ongoing struggles.

Empowering Developers with On-Device AI

Apple’s decision to grant developers access to its Apple Intelligence framework marks a pivotal moment in its AI journey. By providing tools to integrate on-device AI models, Apple is empowering developers to create smarter, more responsive applications that operate directly on users’ devices. These models, boasting around 3 billion parameters, are optimized for efficiency, leveraging the raw power of Apple’s custom silicon to deliver seamless performance without relying on cloud infrastructure.

This approach offers several advantages. On-device processing ensures faster response times, enhanced privacy, and reduced dependency on internet connectivity—a hallmark of Apple’s commitment to user security. Developers can now build apps that perform tasks like natural language processing, image recognition, and personalized recommendations entirely on-device, creating a more fluid and private user experience.(Source: Apple newsroom)

However, there’s a catch. The 3-billion-parameter models, while impressive for on-device applications, are less sophisticated than the larger, cloud-based models powering competitors’ AI systems. This limitation confines developers to less complex tasks, such as basic text analysis or localized machine learning, rather than the advanced reasoning capabilities of cloud-driven AI.  For developers accustomed to the flexibility of cloud-based APIs, this constraint may feel like a double-edged sword: a powerful tool, but one with a narrower scope.

A Step Toward an AI-Driven Ecosystem

The WWDC announcement signals Apple’s ambition to weave AI deeper into its ecosystem, fostering a new generation of intelligent apps tailored to its hardware. Imagine a fitness app that adapts workouts in real-time based on a user’s heart rate, or a note-taking tool that organizes thoughts using on-device natural language understanding—all without sending sensitive data to the cloud. By opening Apple Intelligence to developers, Apple is laying the groundwork for such innovations, potentially redefining how users interact with their devices.

This move also reflects Apple’s broader vision of AI as a seamless, invisible assistant. Unlike competitors who emphasize flashy, all-purpose AI chatbots, Apple is doubling down on integrating AI into everyday tasks. The focus on on-device processing aligns with this philosophy, ensuring that AI feels like a natural extension of the device rather than a standalone feature.

Siri’s Struggles Cast a Shadow

Despite the excitement surrounding Apple Intelligence, the conference wasn’t without its challenges. Apple’s AI strategy has faced intense scrutiny, particularly regarding Siri, its virtual assistant. Reports suggest that Siri’s AI upgrades, long promised as a cornerstone of Apple’s intelligent ecosystem, continue to lag behind expectations. While competitors like Google and Amazon have pushed their assistants to handle complex queries and contextual conversations, Siri remains criticized for its limited understanding and inconsistent performance.
(Multiple sources, including The Verge and WSJ, confirm Apple delayed Siri’s full AI revamp to 2026, citing internal development hurdles.)

For many, Siri’s underperformance is a sore point. Users have shared anecdotes in social media platforms of asking Siri for simple tasks—like setting a reminder or finding a nearby restaurant—only to receive irrelevant responses or errors. At WWDC 2025, Apple acknowledged these shortcomings, hinting at future improvements but offering few concrete details.

This vagueness has fueled skepticism about whether Apple can catch up in the AI race, especially as rivals continue to set the pace with more advanced language models.

Navigating the Competitive Landscape

Apple’s cautious approach to AI—prioritizing privacy and on-device processing—sets it apart from competitors who rely heavily on cloud-based systems. While this strategy resonates with Apple’s privacy-first ethos, it also raises questions about its ability to compete in a market where scale often dictates capability.

Companies like OpenAI and Google leverage massive cloud-based models to power their AI, enabling feats like real-time language translation or intricate problem-solving. OpenAI’s GPT-4 is estimated to use over 1 trillion parameters via a Mixture of Experts architecture, while Google’s Gemini 2.5 models, though undisclosed in size, operate in a reasoning-first, multi-modal cloud environment with advanced capabilities across text, code, audio, and images.

Apple’s 3-billion-parameter models, though efficient and privacy-focused, may struggle to match these capabilities in the short term.

Still, Apple’s track record suggests it’s playing the long game. By focusing on developer access and on-device AI, the company is building a foundation for a sustainable, privacy-centric AI ecosystem. If executed well, this could position Apple as a leader in practical, user-focused AI applications, even if it doesn’t dominate the headlines with the largest models.

What’s Next for Apple Intelligence?

The WWDC 2025 announcement is just the beginning. Apple has promised to roll out additional tools and documentation to support developers, alongside updates to its AI framework in iOS, iPadOS, and macOS. The company also teased enhancements to its machine learning APIs, which could simplify the integration of AI into apps.

For users, the impact of these changes will unfold gradually. As developers experiment with Apple Intelligence, we can expect a wave of innovative apps that make devices feel smarter and more intuitive. Whether it’s a productivity tool that anticipates your needs or a creative app that transforms your ideas, the potential is vast.

Yet, the success of Apple’s AI strategy hinges on addressing its weaknesses. Improving Siri’s capabilities and bridging the gap between on-device and cloud-based AI will be critical. Apple’s ability to balance its privacy principles with the demands of modern AI will determine whether it can reclaim its reputation as an innovator in this space.

At its core, Apple’s WWDC 2025 announcement is about more than technology—it’s about empowering people. By giving developers the tools to create AI-driven apps, Apple is inviting a global community to reimagine what’s possible. For the average user, this could mean a device that feels less like a gadget and more like a trusted companion, anticipating needs and simplifying tasks in ways that feel almost magical.

The promise of smarter, privacy-focused apps is thrilling, but the road ahead won’t be easy. Siri’s struggles remind us that even a tech giant like Apple faces challenges in mastering AI. Still, if there’s one thing Apple has proven time and again, it’s the ability to turn ambition into reality—one carefully crafted update at a time. For now, WWDC 2025 has set the stage for a new chapter in Apple’s AI story. As developers dive in and users await the results, one thing is clear: the future of Apple Intelligence is full of possibilities, and the world is watching.

Also read: Apple’s AI Glasses: Launch Expected in 2026

Workout Buddy on Apple Watch: AI Fitness Coach in watchOS 26