Apple is once again trying to reimagine Siri. The company has built an internal chatbot, codenamed Veritas, that looks and feels a lot like ChatGPT. Employees can use it to ask questions, carry on extended conversations, and even test new features like searching through personal data or editing photos with voice commands. The tool isn’t intended for public release, at least not yet. Instead, it’s a proving ground for a revamped Siri that Apple now says could arrive in early 2026, years after its rivals leapt ahead in the AI race.
The delay is Apple’s central challenge: how can they deliver an assistant that is more than a gimmick without undermining its reputation for reliability? When OpenAI and Google pushed generative AI tools into the hands of millions, Apple largely stayed on the sidelines. Its big AI initiative, called Apple Intelligence, was announced with fanfare in 2024, but the most anticipated part (the smarter Siri) never materialized. For Apple, the promise of a Siri that could understand context, act across apps, and handle natural back-and-forth conversations remains just that: a promise.
That gap is beginning to carry costs. In March, Apple was hit with a consolidated class-action lawsuit alleging that it misled consumers by advertising new Siri capabilities with the iPhone 16 launch that were not actually ready. Plaintiffs argue that Apple used AI hype to drive upgrades and then quietly walked back its commitments, pointing to ads that featured an enhanced Siri, ads the company later pulled. Apple has asked the court to dismiss the case, arguing that feature rollouts were always framed as phased and that customers have already received dozens of other Apple Intelligence tools. Still, the lawsuit highlights how much Apple’s handling of AI has chipped away at its reputation for only announcing features when they’re truly ready.
