Apple is finally entering the field of generative artificial intelligence with the help of an unlikely partner: OpenAI.
Apple CEO Tim Cook today announced the company’s long-awaited AI refresh at the company’s Worldwide Developers Conference in Cupertino, Calif. This fall beta, which the company is calling “Apple Intelligence,” includes several features that will shape the iOS experience in ways big and small. Apple also gave its currently limited-feature voice assistant, Siri, a major generative AI overhaul.
Apple also announced that it will incorporate external AI models into its software, starting with OpenAI’s ChatGPT later this year, and clarified that the experience is opt-in only and does not require a ChatGPT subscription. Siri will determine if a query would benefit from accessing ChatGPT and will ask for approval to share information with the model. Apple said it will also leverage other AI models in the future.
Until now, Apple has been conspicuous by its absence from the headlines regarding generative AI, as competitors such as Google and Microsoft have rushed to adopt the technology since ChatGPT emerged in late 2022. Apple has apparently been biding its time, developing a comprehensive strategy that touches on many of its products and services.
Apple’s senior vice president of software engineering, Craig Federighi, said the company’s new AI strategy will be focused on privacy and security. In addition to highlighting privacy-conscious, on-device AI use cases, Federighi introduced what the company calls private cloud computing, a technology it claims will protect data even for more intensive AI tasks that require the cloud. “User data is never stored or accessed by Apple,” he said.
Apple showed off Apple Intelligence in a number of apps. The technology can rewrite messages in Mail, generate new emoji on demand, and summarize content in Safari. AI updates to Siri help the assistant better handle complex voice commands and find information even when the command is ambiguous.
Apple will also offer smaller generative AI algorithms that run on the device, which the company says will be faster and better protect user data. The company will have to strike a balance between prioritizing privacy and security and pivoting to more generative AI.
Leading up to WWDC, rumors were circulating that Apple was working on a deal with OpenAI. At first glance, Apple and OpenAI seem like a particularly unlikely pairing. Apple is careful and cautious when launching new products, with a focus on attention to detail, attention to user experience, and compelling design. OpenAI is best known for pushing powerful but experimental new AI technologies out into the world and watching what people do with them.
But a deal with OpenAI could make a lot of sense for Apple, which is trying to accelerate its use of large-scale language models. Allowing Siri to use OpenAI’s chatbots gives users access to the most advanced service on the market, while reducing Apple’s own reputational risk if it falters on the most difficult queries. The strategy also buys Apple time to catch up with what OpenAI can currently offer in its AI models.
“With Apple’s announcement, all the big tech companies have laid out their early strategies for generative AI,” says Anyi Wu, an associate professor at Harvard Business School who studies tech companies’ use of AI. Despite being a relatively late entrant, Apple has a big opportunity because of its huge user base, he says.
But Wu says the cost of running a powerful generative AI model and managing its tendency to falsify information would pose new challenges for the company. “To deploy this technology today would require taking those risks, and doing so would be at odds with Apple’s traditional inclination to deliver sophisticated products over which it has complete control,” he says.
This is a developing story, so check back for the latest updates or follow our WWDC live blog here.
