Apple executives say new AI-powered Siri wasn’t ‘demoware’ — it just wasn’t ready to ship

In a handful of interviews following Apple’s Worldwide Developers Conference (WWDC 25), Apple executives denied that last year’s demonstrations of a personalized, AI-powered Siri were vaporware, despite having yet to launch.

When asked by The Wall Street Journal why Apple, with all its engineers and cash, couldn’t make the technology work well enough to ship, the company didn’t admit to being behind in the AI race. Instead, Apple senior vice president of software engineering Craig Federighi stressed that AI was a new technology, and something Apple sees more as a “long-term transformational wave” that will impact the industry and society for decades to come.

“There’s no need to rush out with the wrong features and the wrong product just to be first,” Federighi noted.

Federighi also explained, in an interview with Tom’s Guide and Techradar, that Apple showed off the new Siri at WWDC 24 because the company knew the world wanted “a really complete picture of what’s Apple thinking about the implications of Apple Intelligence and where it’s going.”

He said that Apple had two versions of the AI architecture for Siri, the first of which (version 1) it demonstrated in the video shown at the event. But as development progressed, the team knew that it would have to move the version 2 architecture if it wanted to meet customers’ expectations. This new version is still set to ship in 2026, Federighi confirmed.

The execs also pushed back against the idea that Apple had not shown off functional technology at WWDC 24.

Federighi told the Journal: “We were filming real working software with a real large language model with real semantic search.”

Apple senior vice president of worldwide marketing Greg Joswiak added, “There’s this narrative out there that it’s demoware only. No, it was … something we thought, as Craig said, we’d actually ship by later in the year.” Joswiak said Apple realized it would disappoint customers if it did so, because the software had an “error rate that we felt was unacceptable.”

The execs also talked more broadly about Apple’s plans for AI, which are not to build a chatbot to rival ChatGPT and others, but to infuse intelligence across its operating systems.

“This wasn’t about us building a chatbot … we weren’t defining what Apple Intelligence was to be our chatbot,” Federighi told Tom’s Guide. “That was never our goal … We want to bring intelligence deeply integrated into the experience of all of our platforms in a way that’s ‘meet you where you are’ — not that you’re going off into some chat experience in order to get things done.”

Apple’s real goal, the execs said, was to give developers tools to tap into Apple’s foundation models to build more intelligent apps.