Apple Contemplates Delegating AI Development for Siri Overhaul to OpenAI or Anthropic

Apple Contemplates Delegating AI Development for Siri Overhaul to OpenAI or Anthropic

Apple Contemplates Delegating AI Development for Siri Overhaul to OpenAI or Anthropic


Apple is presently in talks with Anthropic and OpenAI to potentially boost Siri’s capabilities by integrating advanced large language models (LLMs). This development could signify a notable change in Apple’s strategy towards artificial intelligence, as the firm contemplates setting aside its proprietary AI models.

As per a report by Bloomberg’s Mark Gurman, Apple has asked both OpenAI and Anthropic to create tailored versions of their LLMs that would run on Apple’s Private Cloud Compute infrastructure. This strategy is intended to minimize dependence on outside cloud services like AWS or Azure, enabling Apple to have enhanced control over user privacy by leveraging its own servers powered by Apple silicon.

### Ongoing Negotiations

The initiative to overhaul Siri is led by Apple executives Mike Rockwell and Craig Federighi, following a decline in trust in prior leadership. While a comprehensive roadmap for a fully Apple-driven LLM Siri by 2026 is still in place, testing has suggested that Anthropic’s technology might currently align better with Siri’s needs than Apple’s existing models. Corporate development VP Adrian Perica is overseeing negotiations with Anthropic.

Despite the consideration of third-party models, no definitive decision has been finalized. Within the company, Apple continues to advance its competing initiative, known as LLM Siri, which employs in-house models.

### Financial Considerations and Internal Morale

Reports indicate that negotiations between Apple and Anthropic have faced hurdles, particularly concerning financial conditions. Anthropic is pursuing a multibillion-dollar annual fee that rises substantially each year, which has led Apple to explore alternatives, including OpenAI, if a consensus cannot be achieved.

Moreover, the potential shift towards external models has influenced morale within Apple’s Foundation Models team. Some team members believe they are being held accountable for the company’s gradual advancements in generative AI, leading to exits from the team. Compensation concerns have surfaced, with rivals like Meta and OpenAI offering significantly higher salary packages, at times reaching $40 million per year. This pay gap has compelled Apple to issue counteroffers to keep crucial talent.

In conclusion, Apple’s discussions with Anthropic and OpenAI represent a key moment for Siri’s future, as the company weighs the advantages of external AI models against its internal development pursuits. The results of these talks will likely influence the course of Apple’s AI strategy in the near future.