Apple is quietly building a future version of Siri powered by large language models (LLMs). According to recent reports, the company has developed an internal ChatGPT-like app that is being used to test the technology before a public rollout. While consumers cannot access this app, insiders say it is shaping what Apple calls the “next generation Siri.”
This marks a major shift for Apple. Siri, once a pioneer in voice assistants, has lagged behind rivals like Google Assistant and OpenAI’s ChatGPT. The new direction signals Apple’s intent to catch up—and perhaps leapfrog competitors—by fusing its signature focus on privacy and design with advanced conversational AI.
What the Internal Siri App Can Do
The secret app, believed to be code-named Veritas, goes far beyond today’s Siri. It is designed to:
Remember past interactions: Unlike current Siri, which treats each query in isolation, the LLM-powered version can recall earlier conversations.
Work across apps: Apple engineers are testing commands like editing photos or handling tasks in other apps directly through Siri.
Tap personal data securely: The system may pull from a user’s emails, music library, or documents to deliver more personalized results.
Act conversationally: Testers describe it as less robotic and more natural, closer to a true dialogue with the user.
If successful, this could make Siri a genuine competitor in the growing world of generative AI assistants.
Why Apple Is Testing Behind Closed Doors
Apple has chosen not to release the app publicly yet. Insiders suggest that while the technology is promising, it has not consistently met Apple’s high standards for accuracy, reliability, and privacy. The company is cautious after earlier attempts to boost Siri failed to deliver the seamless experience executives demanded.
By keeping development internal, Apple can refine features, test performance at scale, and ensure its on-device privacy safeguards are strong before any consumer launch.
Expected Launch Timeline
Apple reportedly plans to introduce the revamped Siri in early 2026, possibly tied to an iOS 26.4 update. The rollout could become one of the biggest software upgrades since the launch of Apple Intelligence earlier this decade.
However, Apple’s track record suggests the timeline could shift if the company feels the technology isn’t fully ready. As with other major innovations, Apple prefers to delay rather than compromise on quality.
Challenges Apple Must Overcome
While the prospect of an LLM-powered Siri is exciting, there are hurdles:
Privacy concerns: Allowing an AI assistant to access emails or personal files will raise questions, even with Apple’s strong on-device focus.
Third-party model use: Apple may blend its own AI with external models from companies like OpenAI or Google, which could complicate branding and control.
User experience: Not all users may want a chatbot-style interaction. Apple must ensure the new Siri feels intuitive and not overwhelming.
Competition: Google, Microsoft, and OpenAI are already entrenched in the generative AI race. Apple must deliver something distinct.
What It Means for Users
If Apple succeeds, the Siri of 2026 could look very different from the assistant we know today. Instead of short commands, users may engage in natural conversations. Tasks that once required tapping through apps—editing photos, writing emails, setting complex reminders—could be handled seamlessly by voice.
For Apple, this would not only modernize Siri but also strengthen its broader ecosystem, making iPhones, iPads, and Macs smarter and more indispensable.
Apple’s internal testing of a ChatGPT-like Siri app highlights a pivotal moment for the company. After years of being criticized for lagging in AI, Apple appears determined to reclaim its place at the cutting edge.
The real question is whether Apple can balance innovation with its core principles of privacy, reliability, and user trust. If it can, the launch of an LLM-powered Siri in 2026 may redefine not just voice assistants but the entire way people interact with their devices.
