Apple Throws Punches at AI Industry While Quietly Borrowing ChatGPT
WWDC 2025 reveals a company talking tough without the tech to back it up
Hey Adopter,
Apple spent the beginning of June taking subtle shots at the AI industry while simultaneously re-stating today that they're integrating ChatGPT into half of their products. The irony wasn't lost on anyone paying attention.
The Privacy Play Gets Aggressive
Tim Cook and crew positioned "Apple Intelligence" as the antidote to reckless AI development. They hammered privacy repeatedly, calling their Private Cloud Compute an "extraordinary step forward for privacy in AI." Translation: everyone else is doing it wrong.
The messaging was clear. While others rush AI features to market, Apple builds responsibly. While competitors harvest your data, Apple protects it. Classic Apple positioning, except this time they're critiquing an industry they're barely participating in.
Where do you see Apple in the AI race?
What Apple Intelligence Actually Does
Strip away the marketing and you get a collection of decent features. Writing tools that work across apps. Visual search that can identify objects through your camera. Live translation built into calls and messages. A fitness coach on Apple Watch that analyzes your workout history.
The standout feature? Visual Intelligence extending to your iPhone screen. Point at anything displayed in any app and search for it across other services. It's genuinely useful, assuming it works as demonstrated.
The Developer Gambit
Apple's opening their on-device language models to developers through a new Foundation Models framework. Apps can now tap into the same AI that powers system features without sending data to cloud servers. No API costs, works offline, respects privacy.
This could be significant. Developers get access to capable models without the usual trade-offs. Apps like Kahoot can generate quizzes from your notes without external servers. AllTrails can suggest hikes while you're camping off-grid.
The ChatGPT Contradiction
Here's where things get interesting. Apple spent the keynote positioning themselves as the responsible AI company, then announced ChatGPT integration throughout their ecosystem. Image Playground can generate pictures using ChatGPT. Xcode gets ChatGPT-powered coding assistance. Visual Intelligence lets you ask ChatGPT about screen content.
They're careful to say users control when data gets shared with ChatGPT. But the contradiction remains obvious. Apple doesn't trust AI companies with your data, except when they do.
Missing: The Model
Apple talks about their "generative models" and "Apple-built models" for translation. They mention on-device processing and efficient architectures. What they don't mention is having anything that competes with GPT-4, Claude, or Gemini.
This matters because Apple's critiquing an industry they're not really competing in. They're building features on top of existing AI capabilities while positioning themselves as the thoughtful alternative. It's smart positioning until someone asks what Apple's foundational AI breakthrough actually is.
The Reasoning Study Context
Apple's timing is notable given their recent research suggesting large language models can't actually reason. The study found that models perform worse when problems include irrelevant information, indicating they rely on pattern matching rather than logical reasoning.
Now they're building features around these same limitations while critiquing the industry for moving too fast. It's either principled restraint or elaborate justification for being behind.
What Actually Matters
The developer access could change things. If Apple's on-device models prove capable enough for real applications, they've created a compelling alternative to cloud-based AI. Privacy-preserving, cost-effective, and always available.
The integration across devices is also meaningful. Live translation that works in calls, messages, and FaceTime. Visual search that extends to any app. Writing tools available everywhere. When it works, the experience feels seamless.
The Real Test
Apple's betting that thoughtful integration beats raw capability. That privacy matters more than having the most powerful models. That developers will choose their framework over alternatives.
They might be right. But it's easier to take the high road when you're not competing directly. Apple's playing a different game, which is smart until that game becomes the only one that matters.
For now, they've positioned themselves as the grown-up in a room full of reckless kids. Whether they can deliver on that promise without relying on those same kids remains to be seen.
Adapt & Create,
Kamil
P.S. Here’s a list of all Apple AI and Apple Intelligence Features touched upon at WWDC 2025
Core Apple Intelligence Foundation
Private Cloud Compute - Extends iPhone privacy to cloud processing
Foundation Models Framework - On-device language models accessible to developers
App Intents - Connect app capabilities to system AI features
On-device processing - AI models that work offline
Multi-language expansion - Apple Intelligence supporting more languages
iOS 26 Apple Intelligence Features
Communication & Translation
Live Translation - Real-time translation in Messages, FaceTime, and Phone calls
Voicemail summaries - AI-generated summaries of voicemails in Phone app
Messages poll suggestions - AI detects when polls might be useful in group chats
Smart Reply - One-tap response generation
Call Screening - AI answers unknown calls and provides caller information
Spam detection - On-device spam filtering for messages
Visual & Creative
Visual Intelligence - Camera-based object and location recognition
Screen Visual Intelligence - Search and take action on any app content
Genmoji - Generate custom emojis from text descriptions or emoji combinations
Image Playground - Create images with Apple Intelligence
ChatGPT integration in Image Playground - Generate images using ChatGPT models
Expression and hairstyle editing - Modify generated images of people
Clean Up - Remove distractions from photos
Spatial scenes - Convert 2D photos to 3D effects using computer vision
Productivity
Writing Tools - Available across all text input areas
Email/notification/note summarization - Quick content digests
Memory movie creation - Generate photo/video compilations from descriptions
Natural language photo/video search - Find media by describing it
Calendar event extraction - AI pulls dates/times/locations from screen content
Wallet & Commerce
Order tracking intelligence - AI identifies and summarizes shipping details from emails
Apple Pay order summaries - Intelligent order tracking beyond Apple Pay purchases
watchOS 26 Apple Intelligence Features
Workout Buddy - AI fitness coach analyzing workout history and real-time data
Live Translation - Available in Messages app on Apple Watch
Context-aware Messages actions - Machine learning suggests relevant actions
Smart Stack improvements - AI predicts useful widgets based on location and routine
macOS Tahoe Apple Intelligence Features
Shortcuts with Apple Intelligence - Intelligent actions for automation
Direct model access in Shortcuts - Tap into AI models for custom workflows
Live Translation - Cross-platform translation capabilities
Image Playground and Genmoji - Creative AI tools on Mac
Neural Engine optimization - Enhanced performance for AI models
iPadOS 26 Apple Intelligence Features
Live Translation - Communication across language barriers
Shortcut supercharging - AI-enhanced automation workflows
All iOS AI features - Full Apple Intelligence suite on iPad
tvOS 26 Apple Intelligence Features
Apple Intelligence integration - Core features adapted for TV interface
New supported languages - Expanded language support
visionOS 26 Apple Intelligence Features
Image Playground updates - Enhanced creative AI tools
New supported languages - Expanded AI language support
Spatial scenes AI algorithm - New AI for creating lifelike 3D from 2D photos
Computational depth processing - AI-enhanced spatial photo conversion
Developer Tools & Frameworks
Xcode AI Enhancements
Generative intelligence integration - AI assistance in code development
Predictive code completion - More accurate AI-powered suggestions
Swift Assist - Natural language interaction with code
ChatGPT integration - Advanced coding model support
Larger context analysis - AI uses more code context for better suggestions
APIs and Frameworks
Foundation Models Framework - Direct access to on-device language models
Image Playground API - Integrate AI image generation into apps
Live Translation API - Add translation to communication apps
App Intents for Visual Intelligence - Connect app search to Visual Intelligence
Background Tasks API - AI processing can continue in background
External AI Integration
ChatGPT in Image Playground - Generate images with OpenAI models
ChatGPT in Visual Intelligence - Ask questions about screen content
ChatGPT in Xcode - Coding assistance and documentation
User-controlled data sharing - Explicit permission required for ChatGPT features
Technical AI Infrastructure
On-device language models - Core processing without cloud dependency
Private Cloud Compute - Secure cloud processing when needed
Neural Engine optimization - Hardware acceleration for AI tasks
Computational depth algorithms - AI for spatial photo processing
Machine learning context analysis - Smart suggestions based on usage patterns
Generative models - Underlying AI technology powering Apple Intelligence
Text-to-speech AI models - Dynamic voice generation for Workout Buddy
Cross-Platform AI Features
Consistent design with Liquid Glass - AI-informed interface design
Universal device intelligence - AI features work across iPhone, iPad, Mac, Apple Watch, Apple TV, Vision Pro
Seamless data sync - AI insights and preferences sync across devices
Contextual awareness - AI understands usage patterns across device ecosystem
To be fair, it's not uncommon to critique an ecosystem without actively participating in it. Sports commentators do this all the time. Literature critics do this all the time. Thought leaders do this all the time. One might even say that the best critics have a broad outside perspective combined with deep insiders knowledge.
The question is, what role does Apple want in the future? That of a player or a critic?
They announced the partnership with ChatGPT last year? If you're going to be a journalist you have to fact check rigorously.