Apple's Odds of Successfully Launching New Siri in 2026 Are Rising

Gene Munster, Brian BakerJuly 1, 2025

Apple’s Odds of Successfully Launching New Siri in 2026 Are Rising

The prevailing narrative around Apple’s AI efforts is that the company may not be moving fast enough to meet its goal of launching a new Siri by next spring. However, Mark Gurman’s recent report suggests Apple is determined to deliver by considering hitting the reset button on Siri by partnering with companies like OpenAI or Anthropic, or potentially acquiring Perplexity. This more aggressive approach is encouraging for AAPL investors, as it signals the company understands what’s at stake and is willing to think outside the box to get there.

Key Takeaways

It’s becoming increasingly clear to Apple that it can’t deliver an AI-powered Siri on its own and will need to rely on third parties for the underlying LLM.

Expanding the relationship with OpenAI or shifting to Anthropic presents Apple with a near term win and long term challenge.

Perplexity is Apple’s best option because it’s both a strategic fit and allows Apple control.

Apple’s In-House Challenge

Over a one day trading period, shares of AAPL traded up 4% (versus a flat Nasdaq) on a Bloomberg report from Mark Gurman that Apple is exploring using OpenAI’s or Anthropic’s LLM to power Siri instead of their own.

It’s no secret that Apple has been playing catch up in AI. The refreshed Siri, initially demoed at WWDC 2024, was expected in spring 2025, but is now delayed until spring 2026. Last week, we wrote Our Testing Found That Siri Is a Search Tool, Not an AI Assistant, in which we said we believe they should take as much time as necessary to get this right. A rushed rollout risks another high-profile miss, something Apple can’t afford in the AI race.

Gurman’s report underscores the difficulties in pulling the new Siri together and is a reminder that closing the gap in AI is much more difficult than the company’s signature “late to the party and become the life of the party” chapters in portable music, smartphones, and content.

Going Deeper with OpenAI or Anthropic

Currently, Apple has integrated ChatGPT into Siri as part of its new Apple Intelligence features, starting with iOS 18. When users ask Siri complex or creative questions it can’t handle on its own, Siri will ask for permission to pass the query to OpenAI’s GPT-4o model.

This current integration doesn’t replace Siri’s foundation but enhances it by selectively routing certain tasks to ChatGPT. In our testing that I mentioned above, we found that ChatGPT significantly improves the quality of Siri responses but adds unacceptable friction between the process of question and answer between the user and the model. The results are that using the Siri + ChatGPT integration is significantly less natural than direct interaction with ChatGPT.

That means Siri needs to get fixed, and Gurman’s reporting that Anthropic is now in conversations with Apple to replace or sit next to OpenAI underscores the “all options are on the table” urgency inside the company. I believe if they decide to abandon their in-house model, it would be a near term positive given the probability that they land a 10-out-of-10 Siri launch increases.

Long term, there is an important unanswered question. If Apple were to rely on an outside LLM, does that change the company’s control of its future? If these bigger models are of comparable intelligence, widely available, and more interchangeable over time, Apple would have the upper hand because they could threaten to switch models to keep third party LLM providers in check. If the models become more deeply integrated into the DNA of Apple Intelligence, making it difficult for the company to make a switch, then OpenAI or Anthropic would have the upper hand. In some ways, we are seeing this dynamic play out today with the evolving relationship between Microsoft and OpenAI.

Perplexity

The third option is Apple acquiring Perplexity. I believe this makes the most sense given its strategic alignment with Apple’s direction and the fact that it would allow the company to maintain full control of its AI product roadmap.

Here’s why Perplexity would be a smart move for Apple:

Built for Integration: If Apple plans to continue building AI experiences on top of existing foundation models (like GPT-4 or Claude), Perplexity offers the ideal layer of tools to do just that. Its infrastructure is designed to work with major models rather than compete against them. This aligns well with Apple’s quiet, layered approach to AI.

Agentic Browsing: Perplexity is working on an “agentic” AI browser called Comet, which is shaping up to be a rival to OpenAI’s Operator. If Apple were to fold this into Safari, it could transform the browser into a more intelligent AI assistant, making web navigation proactive, not just reactive.

Reimagining Search: Perplexity’s strengths in AI search could help Apple develop a real alternative to Google Search—especially useful with the FTC potentially threatening Apple’s current multi-billion dollar default search deal with Google. This could fast-track a seamless, AI-native search experience deeply tied into the OS across all Apple devices.

Gives more control of their AI future: Perplexity would make it easier for Apple to switch between GPT, Claude, Grok, and Gemini given its library of LLM integration tools.

Disclaimer