Apple turns to OpenAI and Anthropic for Siri, considers abandoning in-house models
Apple wants to use custom models of both companies’ flagship AI offerings so that everything can still stay private. This way, the company can continue to market Apple Intelligence as a service that values user data privacy.
Though a new Siri is still under research and development, the project will likely be scrapped if this agreement goes through. ChatGPT already helps out on Apple devices, but Siri itself still uses the company’s own models.
Apple Intelligence features promised back in WWDC 2024. | Video credit — Apple
Such an approach would be very similar to what Samsung has done: using Google’s Gemini for Galaxy AI on its devices. The Galaxy S25 series is leagues ahead of the iPhone 16 when it comes to AI, but a new Siri powered by either ChatGPT or Claude would catch up almost immediately.
If this goes through, we may even see the new Siri that Apple has promised before its currently expected release of 2026, though I wouldn’t hold my breath given past broken promises. Apple has usually been fashionably late to emerging technologies, but AI just isn’t a field where it can afford to follow its code of conduct.
New product categories are taking shape: for example, AI-powered smart AR glasses. If Apple doesn’t catch up in AI, then it will be way too late to the market with these new products. And, considering CEO Tim Cook’s new obsession, I doubt that that’s something that the company is willing to let happen.
