“Apple Intelligence,” the company’s answer to the likes of Google Gemini and Microsoft Copilot, didn’t exhibit many tricks not already seen on those other platforms, but it did excel in two areas: integration and privacy.
The technology, announced in a prerecorded presentation Monday at Apple’s annual World Wide Developers Conference, largely taps into a user’s personal information to perform its AI functions and does much of it on the devices where the data is stored to ensure privacy.
“There are already some really impressive chat tools out there that perform a vast array of tasks using world knowledge, but these tools know very little about you or your needs,” Apple Vice President of Software Engineering Craig Federighi said in the presentation. “With iOS 18, iPadOS 18, and macOS Sequoia, we are embarking on a new journey to bring you intelligence that understands you.”
He noted that Apple Intelligence can be used to understand and create language, as well as images, and take action to simplify interactions across devices and apps. It can do things like prioritize notifications and provide writing tools to rewrite, summarize, and proofread text.
“They’re deeply integrating AI into your personal data and enhancing that data using artificial intelligence to make your apps smarter, integrate features, and expand capabilities,” explained Tim Bajarin, president of Creative Strategies, a technology advisory firm, in San Jose, Calif.
“It’s a unique approach,” he told TechNewsWorld. “Apple is the only one who can do it because they control the hardware, software and semiconductors in their system.”
A Personal Kind of AI
Mark N. Vena, president and principal analyst at SmartTech Research in Las Vegas, added that Apple’s vision of Apple Intelligence is based on personalized experience.
“People have been focusing on ChatGPT and other AI applications for publicly available content,” he told TechNewsWorld. “Apple’s view is that the real value of Apple Intelligence is how users will use it in a very personalized and customized way.”
“Apple Intelligence will offer a number of interesting applications that the average person will be able to leverage and see the value of,” he said.
“They’ve thrown the gauntlet down,” he declared. “They’re saying that if you really want to drive AI for the common person, it has to be personalized, it has to be customizable, it has to have meaningful applications, and it has to have a privacy component built into it.”
Private Cloud Compute
The cornerstone of Apple’s personal intelligence system is on-device processing, Federighi explained. It allows Apple Intelligence to be aware of personal data without collecting it.
Not all processing can be performed locally, however. Some tasks may need to be performed in the cloud. For those tasks, Apple has created “Private Cloud Compute,” which allows Apple devices to connect to specialized servers running on Apple silicon for enhanced processing.
“These Apple silicon servers offer the privacy and security of your iPhone from the silicon on up, draw on the security properties of the Swift programming language, and run software with transparency built in,” Federighi elaborated.
“When you make a request, Apple Intelligence analyzes whether it can be processed on device,” he continued. “If it needs greater computational capacity, it can draw on Private Cloud Compute and send only the data that’s relevant to your task to be processed on Apple silicon servers. Your data is never stored or made accessible to Apple.”
“And, just like your iPhone, independent experts can inspect the code that runs on these servers to verify this privacy promise,” he said.
“If Private Cloud Compute is true to its word, it’s a nice effort and one that differentiates Apple pretty well,” Eric Abbruzzese, a research director at ABI Research, a technology advisory company headquartered in Oyster Bay, N.Y., told TechNewsWorld.
Vena was impressed with Apple’s care in building the privacy foundation for Apple Intelligence during the presentation. “If you look at all the usage models that they demonstrated, they have to have access to emails and text messages, so it was important to convince people that privacy is at the core of the way they’re approaching AI,” he said.
However, not all AI requests go to Apple servers. OpenAI’s ChatGPT is also in the mix. Siri might determine that a query might be best answered by ChatGPT rather than personal sources. In that case, Siri would ask permission to send the query to the OpenAI chatbot.
“Apple got OpenAI to agree to not log users’ requests, and no personal data can be given to ChatGPT unless you give it permission,” Bajarin explained. “It’s a great approach and extremely secure.”
AI on Apple’s Terms
Ross Rubin, the principal analyst with Reticle Research, a consumer technology advisory firm in New York City, noted that Apple Intelligence offers the company an opportunity to leverage the power of its processors to do something besides rendering video or playing a game faster.
“It allows them to bring more of these AI models onto the device, which helps their privacy stance,” he told TechNewsWorld. “While Apple Intelligence is interacting with a lot of personal information, none of that information is going anywhere.”
“This is AI on Apple’s terms,” he said.
“No one is going to say we’ve never seen AI do this before,” he added. “It’s the way they’ve integrated it into their apps that represents a more comprehensive approach that we’ve seen thus far.”
Anshel Sag, a senior analyst with Moor Insights & Strategy, a technology analyst and advisory firm based in Austin, Texas, agreed.
“They’re not doing anything particularly novel, but their integration at the platform level is good and compelling,” he told TechNewsWorld. “And they’re telling a strong privacy and computer story, which is in line with everything expected from Apple.”
“What they announced is similar to what we’re seeing from other companies, just not at this scale,” he said. “What you’re seeing with Apple Intelligence is a tighter, cleaner integration of what everyone else has already done with a slight Apple twist to it. It’s not particularly groundbreaking.”
More Than Just a Value-Add
Abbruzzese noted that Apple has approached AI as it has approached other products.
“They’re never first to market. They try to be the best,” he said. “It’s too early to know if they’re the best, but I was struck by how cohesive the announcement was. Everything fit together nicely.”
“I haven’t felt the same way for other players,” he continued. “AI always felt like a value-add to something else. It was not as well-integrated.”
“Every AI feature Apple announced works across the ecosystem,” he added. “We haven’t seen as broad-reaching synergy as that before. Microsoft Copilot is powerful, but it doesn’t feel as well integrated into the Windows ecosystem as Apple Intelligence.”