**Apple** unveiled iOS 27 with a new Siri AI on September 14, but analysts and reviewers say Apple is still behind in the AI race compared with Google, OpenAI, and Anthropic.
Key point
Apple releases an iOS 27 beta featuring a newly designed Siri AI on September 14
Requires a separate application and waitlist through Settings; currently limited to English and daily usage
Market assessments continue to say that despite the upgrade, it remains “behind leading AI research labs”
Siri AI launch status
Apple has begun rolling out integration updates alongside iOS 27 on Monday for iPad, Mac, Apple Watch, and Vision Pro. The centerpiece of this update is the extensively overhauled intelligent assistant, “Siri AI,” though the service still has the beta (trial) tag attached.
Users must apply directly for beta participation in the Settings menu, and some users must be placed on a waitlist before receiving access.
Currently, Siri AI runs only in English, while support for French, Japanese, Korean, Portuguese, and Spanish is scheduled for October. A mainland China launch has been put on hold. In the European Union (EU), it is not provided to iPhone, iPad, or Apple Watch users, but it is available to Mac and Vision Pro users.
A daily usage cap applies to cloud-based features, including Siri AI. Apple said it plans to offer “higher usage limits for a fee” in the future, but did not specify the price or when it will take effect.
Related article: XRP futures hit a 6-month high… price diverges as utility growth accelerates
Apple’s AI gap
One cause of the gap is Apple’s approach to AI development. Instead of so-called frontier (hyperscale) models, Apple prioritizes relatively smaller in-house models that can run directly on its own chips, and sends more complex requests to servers running Private Cloud Compute based on Google Cloud. Those servers use **Nvidia** and **Intel** hardware.
**Federico Viticci**, who has reviewed Apple products for years, pointed out that even though Apple’s base models were trained with help from Google’s Gemini, they still fall short of OpenAI’s GPT-5.5 Instant.
Reviewers who have tried the beta version directly say that Siri has finally been able to understand personalized context and leverage a wide range of world knowledge. These are the features Apple promised in 2024. However, at the time of release, market attention had already shifted to agent-type AI—an approach where “you write code, directly control the computer yourself, and handle multiple steps on your own.” As a result, there are quite a few reactions that the overall offering falls short of expectations.
**Zacks Investment Management**’s **Brian Mulberry** described AI as the “biggest single challenge” facing **John Ternus**, who became CEO after Tim Cook on September 1. He emphasized that Apple must prove an “AI vision beyond Siri.”
Other analysts also expect that if Apple successfully reinvents Siri by the end of the year, it could catch up to “a comparable level” to Google’s Gemini-based assistant.
At the 2024 Worldwide Developers Conference (WWDC), Apple unveiled a more personalized Siri for the first time and even ran ads, but the release of its core features was postponed twice—first to 2025 and then to this year. The company reportedly reached a settlement in a class-action lawsuit filed over delays, agreeing to pay $250 million in May. Earlier, in January, it entered a multi-year deal with Google to develop the next-generation model and leverage Gemini technology.
Next read: Iran sells oil, Chinese companies receive cryptocurrency, and now the U.S. Department of Justice goes after $61 million