Apple is reportedly testing Siri’s long-delayed upgrade with the help of an internal ChatGPT-style app. According to Bloomberg, the tool—codenamed Veritas—is being used by employees to trial new AI-driven capabilities and gauge whether a chatbot-style interface has value inside Apple’s ecosystem.
The app won’t reach the public, but it gives Apple engineers a sandbox to push Siri beyond basic queries. That includes retrieving and acting on personal data stored on an iPhone—like searching emails or texts—as well as performing tasks inside apps, such as editing photos. The internal tool also provides a faster way to collect feedback as Apple inches toward a public release of its improved voice assistant.
Siri’s AI overhaul has been years in the making. At WWDC 2024, Apple teased “Apple Intelligence,” a set of AI features meant to balance usefulness with privacy. But when the first wave shipped, the features were modest—functional, yet far short of the personalized, context-aware Siri that Apple had previewed. The flagship update was delayed earlier this year, fueling doubts about the company’s ability to deliver a competitive assistant in a market already crowded by OpenAI, Google, and Anthropic.
The renewed push, Bloomberg notes, suggests Apple is preparing for a 2026 rollout. The company is blending its own AI models with at least one third-party partner. Earlier this summer, Apple weighed integrating models from OpenAI or Anthropic, but recent reports suggest it is now leaning toward Google.
If Apple succeeds, the updated Siri would be far more proactive and capable—closer to the “do-it-for-you” assistant Apple has promised for nearly a decade. But after multiple delays and the lukewarm debut of Apple Intelligence, expectations remain cautious.
