Apple is studying the possibility of using the artificial intelligence technology of Anthropic or Openai for a new version of Siri, instead of relying on its own internal models, reports Bloomberg News on Monday.
The manufacturer of the iPhone would have initiated discussions with the two companies in order to integrate their major language models (LLM) in Siri. According to sources close to the file cited by Bloomberg, Apple would have asked Anthropic and Openai to cause versions of their models that can operate on Apple’s cloud infrastructure, as part of tests.
Still according to Bloomberg, the exploration of third -party models by Apple is only in its infancy and no final decision has yet been made as to their use.
Anthropic, supported by Amazon, declined any comments, while Apple and Openai did not respond to Reuters requests.
Last March, Apple announced that the AI ​​improvements of its Siri vocal assistant would be postponed to 2026, without providing justification to this delay.
To relaunch its efforts in artificial intelligence after several months late, Apple has reworked its management team, entrusting Siri’s supervision to Mike Rockwell. According to Bloomberg, this change follows the loss of confidence of the CEO Tim Cook towards John Giannandrea, then responsible for AI, as for its ability to carry out product development.
In a context of fierce competition between technological giants to win on the flourishing market of generative AI, Apple multiplies partnerships with renowned AI companies and incorporates many AI features directly on its devices to strengthen its offers.
In May, Bloomberg also revealed that Apple was collaborating with Anthropic on a new software platform called “Vibe-Coding”, intended to use AI to write, edit and test code on behalf of programmers.
