J’ai testé le nouveau Siri AI depuis la France, et on rate vraiment quelque chose

J’ai testé le nouveau Siri AI depuis la France, et on rate vraiment quelque chose

I tested the new Siri AI from France, and we are really missing out on something

Title history
  • J’ai testé le nouveau Siri AI depuis la France, et on rate vraiment quelque chose — observed on 20261117-1940 UTC (YouTube)
🎙 Frandroid 👥 626K 📅 November 17, 2026 ⏱ 13 min 👁 2K 📄 expert opinion 🧭 2026-11-17
Available in: English (current) Français

Keywords

Siri AIApple IntelligenceEuropean UnionDMAon-device AI

Summary

The video presents a hands-on test of the new Siri AI, which is currently unavailable on iPhone and iPad in the European Union due to regulatory tensions between Apple and the European Commission regarding the Digital Markets Act (DMA). The presenter, an iPhone user for eight years, managed to bypass the block using a US Apple ID and tested various features: personal context queries (finding messages, notes, contacts), multi-step requests (combining calendar, reminders, notes), and online searches. Most personal context queries worked flawlessly, but online search capabilities were inconsistent, failing to retrieve specific product comparisons or real-time information like social movements. The video then explains the technical architecture: five Apple Foundation Models, with two running on-device and three in the cloud (including one running on Google Cloud with Nvidia GPUs). The core regulatory conflict revolves around the semantic index—an encrypted local index of personal data that Apple argues cannot be opened to third-party assistants without compromising privacy. The DMA requires interoperability, but Apple has refused to launch Siri AI in the EU until a solution is found. The presenter notes that Siri AI is already available on Mac in France, as macOS is not designated as a core platform under the DMA. Potential workarounds (using a US Apple ID) are described but strongly discouraged due to side effects (loss of bank apps, iCloud issues, etc.). The verdict is that while Siri AI is a major improvement, its reliance on Apple-native apps limits its utility in Europe where WhatsApp and Gmail are dominant. The presenter concludes by saying they might switch to a third-party assistant like Claude once alternatives are allowed under DMA, and notes that Apple has already excluded the HomeView device from Europe due to Siri AI, a first for hardware. The video ends with pessimism about Apple’s pace, noting that competitors are advancing rapidly.

306 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video provides valuable practical insights through a real-world test of Siri AI, demonstrating capabilities and limitations with concrete examples. The argumentation is solid: the presenter explains the technical underpinnings (semantic index, five-model architecture) and the regulatory conflict clearly, while acknowledging Apple’s security stance and the EU’s competition concerns. However, the evaluation is based on a single user’s experience without control group, and the presenter’s bias towards switching to Claude is stated rather than justified. The test only covers English-language queries (since US beta), which may not reflect performance in French.

Scientific Rigor, Source Quality, Title Accuracy

The video cites no academic sources but references a Bloomberg report by Mark Gurman and Frandroid’s own articles. The explanation of the DMA conflict is accurate and balanced, presenting both Apple’s and EU’s positions. The title is accurate. The transparency note confirms no sponsorship, and the video explicitly states it was not sponsored. However, the presenter’s personal opinions are clearly distinguished from factual claims, and the regulatory description aligns with known facts about the DMA and Apple’s position.

184 words

Title / Content Match

The title accurately reflects the content: the presenter tests Siri AI from France and concludes that European users are missing out on significant improvements.

Quality & Reliability

7/10

The video provides a hands-on test of the new Siri AI, with clear explanations of the technical architecture and the regulatory context. The presenter openly acknowledges limitations (e.g., online search failures) and distinguishes between personal experience and broader statements. However, the evaluation is inherently subjective and lacks controlled benchmarks or comparative data from other users.

Chapters

Cited Sources

Concurring Sources

  • Apple press release on Apple Intelligence — Context: Official announcement of Apple Intelligence features, though not directly cited in the video, it aligns with the described capabilities.

External References

Contribution & Novelties

The video offers an original hands-on test of Siri AI from a European perspective, detailing the workaround (US Apple ID) and the technical architecture (Apple Foundation Models, semantic index). It highlights the real-world gap between native app support and European usage patterns. The explanation of the DMA conflict is clear and up-to-date.

Pour aller plus loin :

  • Digital Markets Act — Official explanation of the EU regulation affecting Apple.
  • Apple Intelligence — Overview of Apple’s AI features.
  • Semantic Index (Apple) — However, the URL is a guessed developer documentation page; may not exist exactly. Use without URL if uncertain. I will not include a guessed URL.
  • Bloomberg Mark Gurman — The video references his reporting on Siri AI delay; but as a general source, not directly cited.

127 words

Radar Profile

The radar profile shows relatively balanced scores, with moderate quantities of information and solid technical level. The quality and reliability are consistent, but the low score on 'quantite_information' reflects the limited scope of the test (single device, specific use cases). The overall shape is not strongly skewed, indicating a well-rounded video.

Reliability 7/10