Apple CEO Ternus Reportedly Pushing to Turn Siri Into an AI Agent

Apple's newly appointed CEO, John Ternus, is reportedly pushing the company internally to accelerate its AI strategy and transform Siri into a full AI agent, according to a report from Bloomberg's Mark Gurman that ZDNET surfaced for its readers. The pressure comes from the top. Ternus is publicly pleased with the Siri AI launch, Gurman reports, but privately he is encouraging Craig Federighi, Apple's SVP of software engineering, to move faster.

The stakes are straightforward. Apple spent two years hyping Siri AI and shipped an improved voice assistant rather than a groundbreaking AI product, and AI-first competitors are now demonstrating that promised agent features can become reality — some quicker than others.

What can Siri AI do today?

The revamped Siri AI is a much better voice assistant with genuinely helpful features. Gurman singled out one capability as a step toward autonomy: Siri AI's ability to automatically fix weak passwords. He considers that behavior already somewhat agentic (meaning the software takes initiative and completes multi-step tasks on the user's behalf without step-by-step instruction). ZDNET notes, however, that AI-generated passwords carry their own issues.

Siri AI is not yet on par with the increasingly autonomous agents from AI-first companies that aim to proactively organize tasks, use the computer on the user's behalf, and write advanced code. Those rival agents themselves deliver a patchwork of capabilities:

• OpenClaw and Anthropic's Claude Code are highly functional, per the report.
• Google's Gemini Spark is not quite there yet.
• Newcomers like OpenAI's Dot and Anthropic's Duet are not competitive either. ZDNET's AI vetter-in-residence David Gewirtz tested both and found them "a little odd" and not "quite ready for prime time."

The issue, ZDNET argues, is time. Apple has developed powerful hardware perfectly suited for running local AI models (models that execute on the device rather than in the cloud), yet it has remained slower than competitors on the AI software front.

Why has Apple's slow pace been a partial advantage?

Apple's delay has, in the eyes of many observers, actually benefited its reputation amid a rising tide of public backlash against AI. Users are frustrated by software bloat from mediocre or unnecessary AI features, especially as hardware prices climb to accommodate them during a memory shortage.

Users also take comfort in Apple's privacy-first approach to developing AI, which has helped the company stand out against other AI products. The contrast with rivals is concrete:

• Meta's Muse agent may be an extreme example of data-hungry AI agents, but its privacy trade-offs are not altogether absent from other personal agents.
• Google's Gemini relies on more of the user's data to be the most effective version of itself, pitting privacy risk against usability.
• ChatGPT's Computer History feature — which builds a timeline of app and website activity and is available only on the Mac — raised privacy concerns when it launched in August. The feature is opt-in and can be turned off at any time, but OpenAI said it does not encrypt the sensitive information the feature sees, leaving that data vulnerable to hackers.

Yet two years of hype followed by a modest release may mean users are now ready for more advanced features from Apple.

What could an agentic Siri actually do?

An agentic Siri would most likely build on what users already like about the new version. ZDNET offers one concrete example: the on-screen awareness feature is currently limited to what is visible on screen in that moment, meaning only the portion of a long email the user has actively scrolled to. An agentic Siri could ideally scroll through the entire email on its own and surface the insights the user is looking for.

In the short term, Ternus' push will likely just deliver on existing promises. Gurman pointed out that Siri AI still lacks advanced in-app actions, a capability Apple initially announced in 2024 that would let third-party apps create plug-ins for and pathways through Siri.

"This functionality would let Siri tap into both in-house and third-party applications with more precision," Gurman wrote. "This past week, as part of the first beta version of iOS 27.3, Apple finally began talking about the feature again."

Why does the hardware race matter?

AI companies are working on their own hardware, and that threatens Apple's traditional moat. "If they can combine that software with AI-native devices designed from the ground up around assistants that see, remember and act, Apple's longstanding advantages could begin to matter less," Gurman wrote.

The competitive dynamics cut both ways. If Siri AI became a genuinely competitive agent, many Apple users who trust the company's ecosystem might stop installing apps like ChatGPT or Claude on their devices. That remains unlikely today given the wide gap between Apple's AI and the consumer products from OpenAI and Anthropic, but it is not off the table as public AI backlash and privacy concerns mount.

Apple also has an opening on security. If the company pushes forward with AI, it could design more secure in-house versions of features like ChatGPT's Computer History, relying on its well-known and trusted commitment to privacy and to protecting user data.

There is a limit to what users will accept. For those who appreciate the subtlety of Siri AI, ZDNET cautions, anything resembling ChatGPT or Claude running on their devices might not land — unless each feature is opt-in.

The next milestone to watch is concrete: whether the in-app actions framework, revived in the first iOS 27.3 beta, ships as a real capability rather than another announcement, and how quickly Ternus' internal push converts into agent features that close the gap with OpenAI and Anthropic.

(Disclosure noted by ZDNET: Ziff Davis, ZDNET's parent company, filed an April 2025 lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.)