Apple Tests New Internal Chatbot ‘Veritas’ for Siri AI Enhancements

Apple is quietly testing a new internal chatbot named Veritas, designed to accelerate development and feedback for the next generation of Siri. Employees use the tool to ask questions, engage in back‑and‑forth dialogue, and revisit prior exchanges, mirroring the experience of consumer‑facing chatbots like ChatGPT and Gemini. While Apple has not announced plans to release Veritas publicly, the company appears focused on refining Siri’s capabilities amid repeated delays and a lukewarm reception to its recent Apple Intelligence rollout. Analysts note that Apple may still lean on Google’s Gemini for broader AI‑powered search functions.

Background

According to reporting from Bloomberg’s Mark Gurman, Apple is conducting an internal test of a new chatbot called Veritas as part of its effort to modernize Siri. The initiative comes at a time when Apple’s attempts to keep pace in the fast‑moving generative‑AI arena have faced public scrutiny. The company’s next‑generation Siri has experienced multiple delays, and the launch of Apple Intelligence generated a modest, rather than enthusiastic, response from users and industry observers.

Purpose of Veritas

Veritas is intended to give Apple engineers and other staff a rapid prototyping environment for Siri‑related features. By allowing employees to type natural‑language requests, engage in conversational exchanges, and revisit earlier dialogs, the platform mimics the workflow of popular consumer chatbots such as OpenAI’s ChatGPT and Google’s Gemini. The internal tool enables quick testing of functions like searching a user’s personal data and performing in‑app actions, for example editing photos, without the need to build fully‑featured consumer versions first.

Testing Mechanics

Within the chatbot, employees can initiate a query, receive a response, and continue the conversation in a back‑and‑forth manner. The system also supports the ability to scroll back through previous interactions to dig deeper into a specific topic, providing a richer context for iterative development. This approach is designed to surface real‑world usage patterns and edge cases early in the development cycle, allowing Apple to fine‑tune Siri’s behavior before any broader rollout.

Release Outlook

At present, Apple does not appear to have a plan to make Veritas available to the general public. Gurman suggests that the company intends to keep the tool strictly internal, a decision that some observers consider a missed opportunity for consumer engagement. Instead, Apple seems poised to rely on external AI solutions, particularly Google’s Gemini, for its broader AI‑powered search ambitions.

Strategic Implications

The creation of Veritas underscores Apple’s recognition that its existing Siri architecture needs a more agile, conversational testing framework. By leveraging an internal chatbot, Apple can accelerate the feedback loop that is critical for generative‑AI products. However, the choice to keep the system internal may also signal a cautious stance toward releasing experimental AI features directly to gauge user response and manage potential risks.

Industry Context

Apple’s internal focus contrasts with the open‑beta strategies employed by many competitors, who often invite public testers to shape their AI offerings. Meanwhile, the continued reliance on Google’s Gemini for search integration suggests a hybrid approach: Apple builds internal capabilities while partnering with established AI providers for certain functions. This blend reflects a broader industry trend where firms balance in‑house innovation with external expertise to stay competitive.