Siri might be getting a brain upgrade and not just any upgrade, but one possibly powered by OpenAI’s advanced language technology. According to a reliable report by Bloomberg’s Mark Gurman, Apple is currently testing large language models (LLMs) from OpenAI, Anthropic, and Google to decide which one can best bring Siri into the modern AI era.
This move comes after delays and behind-the-scenes changes in Apple’s AI leadership, all aiming to give Siri the kind of smart, helpful personality users expect in 2025.
What’s Happening With Siri?
Siri’s been due for a major update. Apple originally planned to launch a new AI-enhanced version, sometimes called “LLM Siri” back in March 2025. But the company hit some roadblocks. CEO Tim Cook reportedly lost confidence in the direction the AI team was heading, so he brought in Mike Rockwell (previously head of Vision Pro) to take charge of Siri’s next chapter.
Now, Rockwell’s team is experimenting with outside AI brains. That includes:
- OpenAI’s ChatGPT
- Anthropic’s Claude
- Google’s Gemini
Each of these tools is being tested on Apple’s secure private cloud. In simple terms, Apple wants to see which AI model is smartest at understanding everyday tasks, like setting reminders or answering questions, the kind of stuff Siri needs to get better at.
Why Is Apple Looking Outside?
Apple has always preferred to build things in-house. But when it comes to large language models, the kind that power modern AI assistants, the company is behind competitors like Google and Microsoft. So instead of rushing out a half-baked Siri, Apple is wisely exploring the best tools available.
From current testing, Anthropic’s Claude model reportedly showed the most promise. But nothing is final yet.
How This Could Change Siri for You
If Apple does end up using OpenAI or another partner’s tech, Siri could soon:
- Understand your voice requests more naturally
- Hold smarter, more helpful conversations
- Handle more complex tasks without you repeating yourself
For users, this means a Siri that finally feels smart, not just scripted. It’s not about flashy tricks, it’s about real usefulness.
And yes, if this all sounds a bit like what Pixel and Samsung phones already offer through Gemini AI, that’s because it is. Apple is catching up, but doing it carefully, making sure privacy, quality, and smooth performance all meet its standards.
What Happened at WWDC 2025?
Many expected LLM Siri to debut at Apple’s Worldwide Developers Conference in June. But Siri was notably missing from the spotlight. Apple’s marketing lead, Greg Joswiak, admitted the upgraded Siri “didn’t hit our quality standard” yet.
Translation: it wasn’t ready.
That honesty, though, is a good sign. Apple is testing thoroughly before releasing anything, and part of that includes deciding whether it should use outside AI like ChatGPT or Claude at all.
Could Apple Buy Its Way Ahead?
Interestingly, Apple is also exploring another path: buying AI companies outright. Bloomberg reported that Apple had talks about acquiring Perplexity, a fast-growing AI search startup. While nothing’s been confirmed, it shows just how serious Apple is about making Siri smarter, whether that’s by building, partnering, or buying.
Siri vs Competitors Comparison
| Feature | Apple Siri (Current) | Google Gemini (Pixel) | Samsung Galaxy AI (Gemini-based) |
|---|---|---|---|
| Smart Conversation | Limited | Advanced | Advanced |
| Multitask Understanding | Basic | Strong | Strong |
| AI-Powered Summarization | Not yet | Yes | Yes |
| On-device Processing | Yes (Private Cloud) | Some on-device | Some on-device |
| Third-party AI Integration | In Testing | Built-in | Built-in |
Final Thoughts
It’s still early days. Apple hasn’t made a final decision yet. Whether Siri’s future brain comes from OpenAI, Anthropic, or its own labs, one thing is clear: a smarter, more helpful Siri is on the way.
And this time, it might actually be fun and useful to talk to your iPhone.
