Do Son 
Apple is reportedly preparing to launch a new, far more advanced version of Siri in 2026, powered by enhanced artificial intelligence capabilities — and its key technological partner now appears to be coming into focus. According to a report from Bloomberg, Apple will adopt a customized version of Google’s Gemini model to drive the core functionalities of the next-generation Siri, paying Google an estimated $1 billion annually in licensing fees for its use.
The report states that while the new Siri will continue to rely on some of Apple’s own on-device models, it will be heavily dependent on Gemini models operating within Apple’s Private Cloud Compute (PCC) infrastructure to handle “summarizer” and “planner” functions.
These two capabilities are said to be central to Siri’s new architecture — enabling it to synthesize information intelligently and determine how best to execute complex, multi-step tasks. Given that Apple’s broader AI vision centers on empowering Siri to act autonomously within apps on behalf of users, Gemini is expected to serve as the critical engine behind that ambition.
Sources suggest that Apple will pay Google around $1 billion per year for this collaboration — a significant sum, though still a fraction of what Google reportedly spends annually to remain the default search engine on Apple devices.
Despite the scale of the partnership, Apple is unlikely to publicize the deal, maintaining its long-standing emphasis on independence and privacy. The company is also said to be working toward eventually replacing Gemini with its own in-house model.
Internally, Apple is developing a cloud-based foundational model with an extraordinary scale of one trillion parameters, which it aims to ready for consumer applications as early as 2026.
The collaboration marks a notable evolution of Apple’s AI strategy. Its Apple Intelligence framework already allows Siri to invoke OpenAI’s ChatGPT for certain requests, and reports from as early as March 2024 suggested that Google’s Gemini would join as an alternative AI engine for Apple’s ecosystem.
However, after Apple announced in 2025 that the rollout of the redesigned Siri would be delayed by a year, its reliance on external technologies became more apparent. By August, reports indicated that Google had been tasked with developing a customized, privacy-hardened version of Gemini capable of running securely within Apple’s own servers.
Before finalizing its partnership with Google, Apple had also reportedly explored similar deep collaborations with Anthropic and OpenAI.
Related Posts:
- Ransomware in Your Summary? New Attack Weaponizes AI Assistants
- Apple Eyes OpenAI & Anthropic: Claude & ChatGPT Models May Power Next-Gen Siri
- A New “Siri” Is Coming: Apple Considers Using Google Gemini to Power Its AI Overhaul
- The Siri Crisis: Internal Testing Reveals Disappointing Performance for Apple’s Next AI
- FCC Reallocates 6GHz Wi-Fi Band to Mobile Carriers: Threatens Next-Gen Wi-Fi Speeds & Innovation
Support Our Threat Intelligence
Find our tech and OS security coverage helpful? Support our work today and unlock a 100% ad-free reading experience!
Buy Me a Coffee
