It’s Groundhog Day for Siri yet again, as Apple plans another in a long series of delays to the Siri AI upgrade, according to a Bloomberg report released on Wednesday.

According to Bloomberg sources from inside Apple, the reinvented Siri voice assistant, including AI features reminiscent of Alexa Plus, has been delayed from the March iPhone iOS 24.6 update to a release later this year, potentially in May, September or later.

A representative from Apple didn’t immediately respond to a request for comment.

Following repeated delays after announcing the advanced Siri in 2024, Apple gave a broad 2026 timeline for releasing the new voice assistant, widely expected to be included in iOS 26.4. Now, Bloomberg reports that the latest internal testing shows Siri still isn’t up to the task and is likely to remain out of reach for consumers for now. 

This current version of Siri uses an architecture dubbed Linwood, combining Apple’s own large language model with technology from Google’s Gemini AI. The mix is expected to add AI tools such as new web searches and image generation, as well as a chatbot-style Siri AI (code-named Campo), initially slated for iOS 27.

What features are holding Super Siri up?

Much of what Bloomberg reports in this latest release is similar to what we heard last year. The new Siri is too slow, struggles with complex commands, and isn’t meshing well with Apple’s own AI models or with services like ChatGPT. But there are a few new rumors here that show a more complete picture:

• Issues with data access: Bloomberg reports that Apple is pulling back on plans to have Siri scan more personal data, such as reviewing your old text messages to find a shared song or podcast. The feature may come later, but for now it’s being delayed — perhaps consumer concerns over AI privacy are playing a role, too.
• App intents: Intents are Apple’s version of letting Siri complete in-app tasks. For example, Apple wants Siri to be able to find a photo, apply a filter, and post it to your socials or message it to a friend, all in one command. This feature doesn’t appear reliable yet.
• Siri gets terse: Bloomberg contacts also report that Siri is acting up when getting especially quick or complex commands, interrupting users before they can finish and requiring a complete restart.

Why is Apple so late to the voice assistant AI?

I’ve already been experimenting with Alexa Plus and Gemini for Home for months, and I like the results so far. Alexa Plus, in particular (free for Prime users and casual chatters, $20 for the complete package), is much more conversational, understands complex commands, and can tap into brand-new third-party app integrations. Gemini for Home, meanwhile, excels at answering multi-step questions in the app and analyzing video footage. That raises the question of why Apple’s own venture into this space has taken so long.

While I can’t read the minds of Apple devs, it’s clear that Apple invested less in LLMs and generative AI than Google and Amazon did at the start, and is now moving very cautiously into this technology. Given that Alexa Plus doesn’t always know what it can or can’t do and Gemini for Home still struggles with voice chats, I can see why Apple may want to wait for a highly polished product.

Since Apple has been delaying these advanced versions of Siri for well over a year already, it’s clear the company has no problem taking its time and finding a version of Siri AI it’s happy with. According to Bloomberg’s report, we could start seeing betas for these new Siri features in iOS 26.5 or iOS 27 as the slow climb to public release continues.