Google’s new Gemini Spark passes tests that echo the demos shown at last month’s I/O event. In a series of live trials, the model pulled data from a user’s Drive, composed an email in Gmail, and addressed the recipient by first name—all without explicit prompts. The results suggest the system can translate staged promises into use.

Apple plans to embed Gemini Spark into the next generation of Siri, a move that could finally satisfy critics who slammed last year’s concept videos for lacking real functionality. The company’s reliance on Google’s proven AI architecture signals a shift toward outsourcing core intelligence to external platforms, potentially speeding up feature rollouts for iPhone users.

Jay Peters, a Verge reporter, challenged Gemini by requesting an email that summarized his 2026 grocery budget. The model located the correct spreadsheet, calculated averages, and drafted a personalized message, even addressing his wife by name without prior knowledge. This level of contextual understanding demonstrates the AI’s readiness to handle everyday tasks, giving Apple a solid foundation for Siri’s promised capabilities.

With Gemini Spark proving its capabilities in live scenarios, Apple’s upcoming Siri release will likely meet user expectations that have long been unmet. The integration confirms that leveraging external AI can bridge gaps in domestic development timelines. For consumers, this means a more responsive voice assistant arriving sooner than the industry had anticipated.