Apple In Talks With Startup To Shrink AI Models For IPhone
Apple is in talks with PrismML, a startup focused on shrinking artificial intelligence models so they can run directly on an iPhone, according to a report from CNBC. PrismML has confirmed that discussions with Apple are taking place, AppleInsider reported.
The talks center on technology designed to reduce the size and computing demands of AI models while preserving performance, a key requirement for running advanced features on mobile devices. PrismML is backed by Khosla Ventures, livemint.com reported, and has positioned its work around making AI models more efficient for on-device use.
Apple has not publicly detailed the substance of the discussions or whether they could lead to a partnership, acquisition, or product integration. The reporting indicates only that talks are underway and that PrismML has acknowledged them.
The possibility of bringing more capable AI models onto the iPhone matters because it points to a path for delivering AI features without relying as heavily on remote data centers. If AI tasks can be handled on the device, companies can potentially improve responsiveness and maintain functionality in situations where network connectivity is limited. It also suggests a strategy centered on optimizing model performance within the constraints of phone hardware, such as battery life and heat.
For Apple, the discussions fit into a broader push across the tech industry to add more AI capabilities to consumer products. Running AI locally can be a differentiator in mobile computing, where speed, reliability, and tight integration with device software are critical. It can also influence how developers build and distribute features that depend on machine learning, particularly if more processing can occur on-device rather than through cloud services.
The reporting comes as Apple-related market commentary continues to focus on the company’s positioning among large technology stocks, with multiple outlets publishing bullish forecasts and analysis tied to Apple’s prospects and product cycles. Those opinion pieces are separate from the PrismML talks, but they underscore the attention on how Apple plans to incorporate AI advances into its flagship devices.
What happens next is likely to remain private unless Apple or PrismML chooses to disclose more, or if future reporting clarifies the outcome of the talks. In the near term, observers will look for any indication of a formal agreement, a product announcement, or confirmation that PrismML’s model-shrinking technology is being tested for deployment on Apple hardware.
Until then, the only verified development is that Apple and PrismML are in discussions about technology aimed at making powerful AI models small and efficient enough to run on an iPhone—an area that could shape how the next generation of mobile features is delivered.
