Pang joined Meta in mid-2025 after leaving Apple, where he focused on AI model infrastructure.

At Meta, Pang oversaw key systems supporting next-generation AI models. His exit comes after months of recruitment by OpenAI, and sources said he left Meta last week.

Meta and OpenAI did not immediately respond to requests for comment.

The development reveals the scale of investment companies are making in AI talent. Reports indicate his compensation at Meta was more than $200 million over multiple years.

His work at Meta was a major one in developing the company’s next-generation AI systems, part of its long-term strategy to compete with firms including OpenAI, Anthropic, Apple, and Google.

For OpenAI, Ruoming Pang strengthens its infrastructure and ability to scale. His expertise will support the training of larger multimodal models and more advanced AI workflows.

The hire occurs as governments strengthen review of AI development. Canada, for example, has recently urged OpenAI to increase safety measures or face regulatory consequences.

Pang led Apple’s Foundation Models team, responsible for training the AI systems powering recent on-device tools like Genmoji and smart notifications.

His departure, confirmed by insiders, follows growing internal challenges at Apple over its reliance on external models from OpenAI and Anthropic to deliver critical Siri upgrades. 

For a company that prides itself on tight vertical control, outsourcing key AI infrastructure to rivals has raised eyebrows internally and beyond.

Meta, meanwhile, is making no secret of its aggressive expansion plans. Pang now joins its elite AI division, Meta Superintelligence Labs (MSL), which has become a magnet for top talent. 

It’s headed by Alexandr Wang, the former CEO of Scale AI, a company Meta recently invested in at a $29 billion valuation. Bloomberg reports that Pang’s compensation at Meta runs into the millions annually.

Pang’s exit comes weeks after his deputy, Tom Gunter, also left Apple, and is part of a pattern of senior AI defections. Mark Zuckerberg has been personally overseeing this high-stakes recruitment wave, pulling in names like Daniel Gross of Safe Superintelligence and former GitHub CEO Nat Friedman. Researchers from OpenAI, Anthropic, and DeepMind have also reportedly crossed over to Meta.

“Apple’s recent AI enhancements have been minimal, and the more sought-after features, like the Siri revamp, are being pushed out to 2026,” said Dipanjan Chatterjee, Forrester VP and principal analyst. “Apple’s strong brand equity has absorbed much of this damage without customer defection, but this equity is being steadily eroded as Apple fails to make good on its AI promises.”

While Apple made a strong hardware showing at WWDC 2025, its AI reveal leaned heavily on partnerships, rather than showcasing any breakthrough from its internal teams. That silence from within is now being read as a red flag.

Zhifeng Chen has now taken over Pang’s team, but oversight is fragmented across software boss Craig Federighi, Siri and Vision Pro head Mike Rockwell, and a diminished role for John Giannandrea. 

Analysts say this reshuffling is unlikely to change Apple’s pace, particularly in a market where rivals are throwing around $10 million to $100 million offers just to attract top talent.

Sam Altman, CEO of OpenAI, in a recent podcast interview with his brother, said, “Meta had been targeting his employees with signing bonuses that went as high as $100 million.”

There were no huge declarations or tools at the WWDC25, just practical tweaks aimed at improving user experience.

From live call translation to on-device smart summaries, what Apple announced was measured and minimal. Siri didn’t get the overhaul many had expected after last year’s promises. 

Instead, Apple focused on more secure system integration and privacy-led enhancements, nothing that screamed innovation.

Craig Federighi, Apple’s senior vice president of Software Engineering, confirmed developers would now be able to access the on-device large language model embedded in Apple Intelligence. 

“We’re opening up access for any app to tap directly into the on-device, large language model at the core of Apple,” he said. That’s important for privacy, but not necessarily performance.

The on-device model itself is limited. It runs on about 3 billion parameters, small when compared to the power of cloud-based systems used by Microsoft or Google. It can’t process large, complex tasks, which means it is limited in areas where true generative AI is beginning to thrive.

Still, Apple’s strategy is to stay local, stay secure and features like call screening, where iPhones can pick up unknown calls, ask why the person is calling, and then transcribe the response before the phone even rings, are clever. 

So is the live call translation that doesn’t require the other caller to own an iPhone. It’s thoughtful tech, but hardly disruptive.

The redesigned operating systems, featuring what Apple calls a “liquid glass” aesthetic, are another example. It looks sleek, but it’s not revolutionary. It’s enabled by improved Apple silicon, and now all OS platforms, from iPhone to Mac, will adopt a consistent naming convention. The move to names instead of numbers is a way to streamline branding.

Image Playground, which now allows users to generate pictures using ChatGPT, was also showcased. Apple says user data won’t be shared with OpenAI unless the user consents. This cautious collaboration highlights how Apple is trying to balance innovation with its longstanding privacy-first ethos.

What’s missing from all of this is clarity on vision. A year ago, Apple spoke of intelligent agents and a new era of AI. That talk has all but disappeared. Analysts are taking note.

“In a moment in which the market questions Apple’s ability to take any sort of lead in the AI space, the announced features felt incremental at best,” said Thomas Monteiro, senior analyst at Investing.com. “It just seems that the clock is ticking faster every day for Apple.”

That ticking clock got louder as OpenAI, during Apple’s WWDC25 event, announced it had hit $10 billion in annualised revenue. One company accelerating into the AI future, the other inching forward.

Even within the developer community, there are questions. Apple’s Foundation Models Framework allows developers to plug into Apple Intelligence, but only the on-device version. The high-powered, cloud-backed models that could have taken these tools to the next level are staying behind closed doors.

Ben Bajarin, CEO of Creative Strategies, said: “You could see Apple’s priority is what they’re doing on the back-end, instead of what they’re doing at the front-end, which most people don’t really care about yet.”

Investors, it seems, agreed. Apple shares fell 1.2% by the end of the day, hardly a collapse, but a sign that the market wasn’t impressed.

If Apple is laying the foundation for bigger things, it’s doing so without noise. This is a deliberate approach we hope will pay off, not leave the company behind competitors.

Developers and tech professionals are looking beyond routine updates, the company is preparing to announce changes across its software platforms, with higher focus on artificial intelligence.

The event, which combines in-person attendance at Apple Park with online programming, is headlined by the theme “On the horizon.”

Apple is also planning a deep redesign of its software, particularly for iOS 19, macOS 16, iPadOS 19, visionOS 3, and watchOS 12.

From what we’ve gathered, users should expect overhauled icons, refined menus, redesigned apps, and simplified navigation, all aligned with the aesthetic of visionOS.

Apple isn’t just putting a new coat of paint on its platforms. The bigger innovation lies in the integration of “Apple Intelligence” features across its ecosystem. 

These include new tools for smarter battery performance and a virtual assistant for health tracking, obvious signs Apple is pushing further into personalised and predictive user experiences.

The WWDC 2025 conference opens on 9 June at 10 a.m. PDT with the traditional keynote, streamed globally on Apple’s platforms, including the Apple TV app and YouTube. Later that day at 1 p.m. PDT, the “Platforms State of the Union” will offer a technical breakdown of the new tools, features, and APIs that developers will work with in the months ahead.

Developers can expect over 100 sessions during the week, covering updates across Swift, machine learning, graphics, and system frameworks.

For the first time, Apple is adding online group sessions to supplement one-on-one appointments with engineers and designers. This is different from past years, stressing the company’s move to make developer engagement more collaborative and flexible.

A highlight of the conference is the Swift Student Challenge, where Apple is spotlighting 50 standout winners with a three-day immersive experience at Apple Park. The goal is to nurture the next generation of coders and creators from the ground up.

Everything will be accessible online via the WWDC25 website, the Apple Developer app, and official Apple channels.

Apple said in its announcement: “WWDC25 is going to be our biggest and most forward-looking yet.” Given the changes coming, that doesn’t appear to be an exaggeration.

Some say they’re clunky, slow, and miss the point entirely. Now, in a rare pivot, Apple is quietly trying to fix things by tapping into something it has long guarded: your device’s data.

No, they’re not harvesting your emails. But they are sending test data to your iPhone or Mac—if you’ve opted in to share analytics. That’s the catch. They call this process differential privacy, but in real terms, here’s what it means: Apple creates large pools of fake emails, analyses their structure, and then compares them to snippets from real user content—without ever collecting that content.

It’s a workaround. A clever one.

The company admitted in a blog post: “To curate a representative set of synthetic emails, we start by creating a large set of synthetic messages on a variety of topics […] We then derive a representation, called an embedding, of each synthetic message that captures some of the key dimensions of the message like language, topic, and length.”

Apple Sued Over Apple Intelligence Delay; Customers Seek Compensation for Missing AI Features

These “embeddings” are sent to devices participating in Apple’s Device Analytics programme. Those devices then silently check them against your real email samples—not to send back your data, but to inform Apple which of its synthetic datasets hit the mark.

I get why they’re doing it. Apple’s current AI features have fallen behind the likes of OpenAI and Google. Its models are struggling because they’re too reliant on data that doesn’t reflect how people actually communicate. Real-world input is messy, nuanced, and unpredictable. The synthetic stuff? Too polished, too generic.

Apple is hoping this new feedback loop will breathe life into tools like Genmoji, Image Playground, Visual Intelligence, and Writing Tools. And while they claim to respect your privacy—“Synthetic data are created to mimic the format and important properties of user data, but do not contain any actual user-generated content”—the success of this hinges entirely on whether enough users say yes to the background data checks.

There’s more. In the same blog post, Apple explained that it’s using this method to refine summaries, writing suggestions, and even memory-based image creation. It all ties into a bigger initiative the company has branded as “Apple Intelligence”, which—let’s be honest—hasn’t exactly inspired confidence yet.

Interestingly, Apple’s recent management shuffle has its own story. In March, the company stripped Siri from long-time executive John Giannandrea. Control was handed instead to Mike Rockwell, the man behind the Vision Pro, and software boss Craig Federighi. They want results, not promises.

But the timeline is still murky. While the changes are expected to appear in beta versions of iOS 18.5 and macOS 15.5, the more significant AI updates—especially for Siri—won’t land until next year.

It’s a wait-and-see game now. Apple’s new approach is smarter, more private, and frankly overdue. But if people don’t opt in—or if the synthetic datasets still fall flat—then Apple may find itself further behind in a path it used to lead.

One of the interesting features of these updates is “Clean Up,” a tool that allows users to effortlessly remove unwanted objects from their photos.

The Clean Up feature, integrated into the native Photos app, provides users with the ability to remove elements from an image without disturbing the overall composition. 

By simply selecting an object or drawing over it, users can remove distractions from their photos. Apple’s system intelligently fills in the background where the object was removed, even accounting for shadows and reflections to ensure a seamless edit.

This addition comes as Apple continues to expand its range of photo editing tools, aiming to offer users more control over their images directly from their devices. 

While similar features, such as Google’s Magic Eraser, have been available on other platforms, Apple’s Clean Up is designed to integrate deeply with the existing Photos app, making it a convenient and powerful tool for iPhone, iPad, and Mac users.

Currently, Clean Up is available only to users in the United States who have the latest iPhone 15 Pro, Pro Max models, or M1-powered devices. 

Other updates in iOS 18.1 include enhancements to Apple Intelligence features like notification summaries, now extended to cover all apps, and improved natural language search in Photos.

With these updates, Apple is enhancing its software ahead of the full release of iOS 18, expected to coincide with the launch of the iPhone 16 in September. However, Clean Up and other new features will be rolled out more widely in October with the iOS 18.1 update.

These updates bring a redesigned Siri that can maintain context over multiple interactions, new smart reply and email summary features in Mail, and enhanced natural language search capabilities in Photos. 

However, features such as a complete change of Siri are scheduled for gradual release, extending into 2025, as reported by Bloomberg.

To access these new features, users must own an iPhone 15 Pro or Pro Max or a device with an Apple Silicon chip. 

After installation, users are required to join a waitlist via the device’s Settings menu, as indicated by @iSWUpdates on X. 

Concurrent with the debut of the 18.1 developer previews, Apple has also launched a second public beta for iOS 18. This update includes enhancements such as expanded RCS support, new CarPlay wallpapers, and the ability to use dark mode widgets while in light mode.

WWDC 2024: Apple Unveils AI-Powered Future with Siri Boost, ChatGPT Integration, But Faces Musk’s Ire

Historically, follow-up software updates like 18.1 have not been released before the public rollout of the initial version, but Apple has moved from its usual practice. 

In releasing these AI updates as an 18.1 beta, Apple aims to identify and resolve possible issues ahead of a wider launch, ensuring a smooth transition for upcoming iPhone 16 users who will receive iOS 18 this fall.

Currently, Apple Intelligence features are available exclusively to developers, with a broader beta release anticipated soon. These Apple fine-tuned features are expected to evolve, giving users a polished experience upon full launch. The early developer beta lacks some functionalities, which are slated for later release.

Among the new tools, Apple’s Writing Tools provide proofreading, grammar checks, and tone adjustments, with options for creating summaries and lists across various apps. 

Siri now has a visual glow that responds to voice interactions, enhancing usability without distraction. Additionally, users can interact with Siri using text prompts, with a new interface that simplifies typing requests. Siri’s improved contextual awareness allows for more seamless follow-up queries, enhancing user interaction.

Mail improvements include the ability to summarise emails and prioritise time-sensitive messages, ensuring users can quickly identify important information. Smart Reply options offer quick responses based on email content, while the Lock Screen now consolidates multiple notifications.

The Photos app benefits from a natural language search feature, enabling users to find specific images or moments more intuitively. Memory Movie creation allows for the easy compilation of thematic slideshows, complete with music and customisable scenes. 

Again, Apple Intelligence introduces transcription features in Notes and other apps, allowing users to convert audio recordings into text with accompanying summaries, making it ideal for lectures and meetings. The Focus Modes now include a “Reduce Interruptions” setting, leveraging AI to filter non-essential notifications.

iOS 18.1 also introduces phone call recording, notifying all participants and storing the recordings in the Notes app, complete with transcripts. 

Safari gains the ability to summarise articles in Reader Mode, while users can access an Apple Intelligence Report in the Settings app, ensuring transparency in data usage.

The rollout of Apple Intelligence is currently restricted to U.S. English users, with availability limited to select regions.