Apple abandons internal AI development, turns to OpenAI and Anthropic for Siri

Apple is abandoning its internal AI development for Siri and instead considering partnerships with OpenAI or Anthropic to power its voice assistant, according to new Bloomberg reporting. This represents a major strategic retreat for one of the world’s largest tech companies, which has faced lawsuits from shareholders and customers over unfulfilled promises about AI-powered Siri features in the iPhone 16.

What you should know: Apple’s AI ambitions have spectacularly failed to materialize, forcing the company to seek outside help for Siri’s long-promised upgrade.

• The iPhone 16, launched in September 2024 for $799, was marketed with promises of “Apple Intelligence” features and an AI-powered Siri that could autonomously complete tasks.
• Instead of delivering on these promises, Apple has repeatedly delayed the Siri AI upgrade, most recently pushing it to 2026 in early June.
• The company now faces multiple lawsuits from shareholders and customers who claim they were misled about the AI capabilities.

The partnership talks: Apple has approached both OpenAI and Anthropic about training versions of their large language models to run on Apple’s cloud platform.

• Anonymous sources told Bloomberg that Anthropic’s Claude is currently the preferred choice among Apple executives, including Siri chief Mike Rockwell.
• The discussions involve having these AI companies customize their models specifically for Apple’s infrastructure, likely for mass adoption testing.

Why this matters: The move signals Apple’s acknowledgment that it cannot keep pace with AI development internally, marking a significant shift for a company known for controlling its technology stack.

• Apple’s retreat comes as the company has invested heavily in AI infrastructure, including a $500 billion commitment to US tech manufacturing over four years.
• The decision positions Apple as the largest tech company to openly question the current AI hype cycle.

The bigger picture: Apple’s AI struggles coincide with the company’s growing skepticism about industry-wide AI claims.

• In June, Apple’s Machine Learning Research lab published a white paper declaring that the broader AI industry was “massively overhyping” the abilities of top AI models.
• The research specifically challenged OpenAI’s claims that its chatbots can “reason,” a key selling point used by CEO Sam Altman.

What this means: Apple’s potential pivot to third-party AI partnerships would be viewed as a major failure by investors and customers, representing the company’s inability to deliver on its AI promises after nine months of delays and legal challenges.