Apple hits record high on AI expansion plans

Apple shares climbed 3.95% to a record high on Wednesday, propelled by developments signaling the company's expanding AI capabilities across devices and geographies.

On Tuesday, CNBC reported that Apple is evaluating technology from PrismML, a Silicon Valley start-up, to shrink large artificial intelligence models so they can run directly on iPhones. PrismML licenses this technology from the California Institute of Technology. According to PrismML CEO Babak Hassibi, Apple is actively testing the tech's performance on its devices. If successful, this could allow iPhone users to access advanced AI model capabilities without relying on cloud servers—and it could help Apple reduce its cloud computing costs significantly if AI applications run on-device instead of remotely.

On the same day Apple's stock hit its high, a second piece of news emerged: China's Cyberspace Administration approved Apple to provide AI services in the country. This regulatory green light opens a major market. Alibaba, the Chinese e-commerce and cloud giant, will integrate its Qwen AI model into Apple Intelligence. Baidu, China's internet search leader, will also collaborate with Apple to develop AI features for its devices. Together, these partnerships mean Apple can offer AI-powered functionality to Chinese users without building the underlying models itself.

These moves reflect Apple's broader AI strategy, which has largely avoided the expensive race to build proprietary AI models. Instead of spending tens and even hundreds of billions of dollars to compete with model makers like OpenAI and Anthropic, or with hyperscalers like Google and Meta Platforms, Apple has opted to partner with AI leaders and integrate their innovations into Apple devices and services. Investors view this cost-effective approach favorably: it lets Apple benefit from AI breakthroughs and user demand without incurring the massive capital expenditures that competitors are undertaking.

Apple has taken a markedly different path in the AI race than most of its tech peers. While companies such as OpenAI, Google, and Meta Platforms have invested heavily in building proprietary large language models and AI infrastructure, Apple has largely stayed out of the model-building competition itself. Instead, the company is licensing and integrating technology from specialized startups and AI leaders—a strategy that allows it to deliver cutting-edge AI features to users without the enormous R&D expense.

The two announcements on Tuesday and Wednesday demonstrate how this approach is starting to pay off. On-device AI execution via PrismML's technology could be particularly valuable: it would reduce Apple's reliance on cloud servers (lowering operational costs) while improving user privacy and responsiveness by keeping inference local. Meanwhile, the China approvals open a major market where Apple can offer AI services without building the underlying models—Alibaba and Baidu handle that work, and Apple integrates their innovations into its ecosystem. This partnership model appears to be working: investors responded by driving the stock to an all-time high, reflecting confidence that Apple can capture AI's benefits without the billions in capital expenditure its competitors are deploying.