Apple is testing third-party AI models to revive its struggling voice assistant.
According to Bloomberg, Apple has asked OpenAI, Anthropic, and Google to run their LLMs like ChatGPT, Claude, and Gemini on Apple’s private cloud.
Why?
Because Apple’s own AI isn’t cutting it.
Siri’s long-awaited AI upgrade, “LLM Siri,” was delayed earlier this year.
Internal testing reportedly showed that Anthropic’s Claude outperformed Apple’s in-house model.
This comes after major internal shakeups:
– Siri’s former chief, John Giannandrea, was replaced
– Vision Pro lead Mike Rockwell is now in charge of AI
– Tim Cook has “lost confidence” in the original Siri team
Apple is also reportedly considering acquiring Perplexity a fast-growing AI search startup already partnered with Motorola.
Meanwhile, competitors are racing ahead:
– Google’s Gemini is built into Pixel and Android
– Samsung is licensing Gemini and in talks with Perplexity
– Siri didn’t even show up at WWDC
Apple’s marketing head admitted it:
“It didn’t hit our quality standard.”
If Apple chooses OpenAI or Anthropic, it would mark a major shift in strategy, breaking with its history of building everything in-house.
This could redefine Siri and set the stage for Apple’s next chapter in AI.
But one thing’s clear:
Apple knows it’s behind.
And now it’s looking for outside help to catch up.
