Even Apple, with its traditional caution and secrecy, has not been able to ignore the speed at which artificial intelligence is evolving. In October 2024, following its announcement at WWDC, the company began rolling out Apple Intelligence, marking its official entry into the AI field. However, the progress has been deliberately slow: new features are being introduced gradually and, so far, are limited to discreet improvements. The anticipated reinvention of Siri has yet to materialize, and the more ambitious updates remain in development. Meanwhile, Apple is internally experimenting with Veritas, a conversational AI application similar to ChatGPT, used exclusively by employees as a testing environment. This project reveals both the direction and the reservations of Cupertino's strategy. Veritas is a conversational app restricted to Apple staff, allowing interaction with an advanced language model through natural commands and automating tasks with system data. Although its interface resembles that of a conventional chatbot, its integration with Apple Intelligence infrastructure sets it apart: it allows editing notes, searching files, organizing reminders, and managing multiple conversations simultaneously. Its goal is not to compete in the chatbot market but to serve as a laboratory to refine the future Siri. Since the introduction of Apple Intelligence at WWDC 2024, new capabilities have been gradually rolling out. In December, Genmoji and Image Playground debuted; in January, the feature was activated by default on devices compatible with iOS 18.3; and recently, real-time translation was added to Messages and FaceTime. However, the key piece—the new AI-powered Siri—remains pending. Recent leaks suggest March 2026 as a possible internal launch date, but Apple remains tight-lipped. This caution stems from a lesson learned: the company wants to avoid the communication errors of WWDC 2024, when the lack of concrete details caused confusion. Therefore, Apple prefers to refine its advancements behind closed doors. Veritas acts as a simulator for real scenarios, allowing the evaluation of the future Siri's behavior in a controlled environment, away from public scrutiny. Apple's strategy also differs from that of its competitors: it does not seek to launch an independent chatbot but to integrate artificial intelligence in an invisible and useful way into the user's daily experience. Veritas does not contradict this vision; on the contrary, it reinforces it as an internal testing ground, not as a preview of a commercial app. Apple bets on functional AI, not on AI as the main character. Thus, while the latest iPhones already incorporate some features of Apple Intelligence, Veritas remains in the shadows. Its internal use is not a sign of failure but a demonstration of Apple's meticulous approach: a more closed process focused on quality control and, therefore, slower. This pace contrasts with the speed of Google and Microsoft, but aligns with Apple's vision of AI's role in its products. The big question is whether Apple can maintain this tempo without falling behind in an increasingly fast-paced race. While Google integrates Gemini into every new Pixel and Microsoft transforms Windows with Copilot, Apple moves forward steadily, albeit more slowly. Veritas may be the rehearsal it needs to make the definitive leap, or perhaps just a reflection of how much still needs to be refined.