Apple is preparing a significant revamp of its Siri voice assistant, aiming for a spring 2026 launch after delaying the planned 2025 release due to technical challenges. The overhaul involves integrating cutting-edge AI technologies to make Siri more conversational, capable, and human-like in handling complex tasks. To accelerate development and enhance capabilities, Apple is exploring partnerships with major AI companies, notably Google and Anthropic. Early talks with Google focus on using its Gemini AI model for boosting Siri’s summarization features, while discussions with Anthropic involve leveraging Claude for planning functions.
This marks a strategic shift for Apple, traditionally known for relying on proprietary in-house AI models, accepting the need to incorporate third-party large language models to catch up with competitors like Google Assistant and Amazon Alexa. The updated Siri will include three core AI components—a planner to understand user inputs, a search system to comb the web and personal data, and a summarizer to deliver coherent responses. While Siri will utilize third-party AI for web content processing, Apple intends to keep personal data processing under tight privacy controls with its own Foundation Models.
The new Siri testing app, internally known as “Veritas,” is currently used by Apple employees for rapid evaluation of these AI features but will not be made publicly available. Apple views this integrated AI upgrade as part of a broader Apple Intelligence strategy to embed smarter, privacy-preserving AI throughout its devices, rather than offering a standalone chatbot product. This approach reflects Apple’s cautious but evolving commitment to maintaining privacy while advancing AI capabilities in its ecosystem.
