Ready for a Siri AI agent? Apple under pressure to upgrade its assistant faster

ZDNET’s key takeaways

• Bloomberg’s Mark Gurman reports Apple may push harder on its AI strategy.
• That would mean a more agentic Siri.
• Apple could lose ground to AI-first rivals.

Apple’s recently appointed CEO, John Ternus, is allegedly pushing the company internally to ramp up its AI strategy — including making Siri an AI agent.

Also: My 5 favorite Siri AI ‘onscreen awareness’ tricks

According to Bloomberg’s Mark Gurman, the CEO is publicly pleased with the Siri AI launch, but privately encouraging Craig Federighi, Apple’s SVP of software engineering, to move faster.

Siri AI and Apple’s struggle

The revamped Siri AI is now a much better voice assistant with some helpful features. Gurman noted he thinks Siri AI’s ability to automatically fix your weak passwords is already somewhat agentic behavior, though AI-generated passwords have their own issues.

But Siri AI isn’t yet on par with increasingly autonomous agents from AI-first companies that aim to proactively organize your tasks, use your computer, and write advanced code. Still, those rival agents deliver a patchwork of capabilities: while some, like OpenClaw and Anthropic’s Claude Code, are highly functional, others, like Google’s Gemini Spark, aren’t quite there yet.

Also: I tried replacing Claude Code with a free local AI stack

Regardless, the issue here is time. Despite developing powerful hardware perfectly suited for running local AI models, Apple has remained slower than competitors on the AI software front. Competitors have demonstrated they can make promised agent features a reality, some quicker than others.

Apple’s slower pace in AI has, in the eyes of many, actually benefited its reputation amid a rising tide of public backlash against AI. Some users are frustrated by the software bloat from mediocre or unnecessary AI features, especially as hardware prices skyrocket to accommodate them during a memory shortage. Users also take comfort in Apple’s privacy-first approach to developing AI, which has helped it stand out against other AI products.

Meta’s Muse agent might be an extreme example of data-hungry AI agents, but its privacy trade-offs aren’t altogether absent from other personal agents. Gemini relies on more of your data to be the most effective version of itself, pitting privacy risk against usability.

Also: Meta Muse is the worst AI agent for privacy

On the other hand, the fact that Apple hyped Siri AI for two years only to release an improved voice assistant (rather than a groundbreaking AI product) may mean users are ready for more advanced features.

In addition, AI companies are working on their own hardware. “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.

What an agentic Siri AI could mean

An agentic Siri would most likely build on what users already like about the new version.

The on-screen awareness feature, for example, is currently limited to what’s visible on your screen in that moment, meaning only the part of a long email that you’ve actively scrolled to. An agentic Siri could ideally scroll all the way through an email on its own and surface the insights you’re looking for.

Also: The iPhone made Apple a giant – but the Mac could help it win the AI era

But in the short term, Ternus’ push will likely just deliver on existing promises. Gurman also pointed out that Siri AI still lacks advanced in-app actions, which Apple initially announced in 2024 and 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.”

If Siri AI were a competitive agent, many Apple users who trust the company’s ecosystem might move away from installing apps like ChatGPT or Claude on their devices. That’s still unlikely as of now given the wide gap between Apple’s AI and consumer products from OpenAI and Anthropic, but not off the table as public AI backlash and privacy concerns mount.

(Disclosure: 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.)

Plus, if we’re talking about agents specifically, newcomers like OpenAI’s Dot and Anthropic’s Duet aren’t competitive themselves yet: ZDNET’s AI vetter-in-residence David Gewirtz found both to be “a little odd” and not “quite ready for prime time”.

Also: I tested OpenAI Dot and Anthropic Duet for weeks – and all I got was one helpful notification

ChatGPT’s Computer History feature — which creates a timeline of your app and website activity, and is only available on the Mac — raised privacy concerns when it launched in August. While the feature is opt-in and can be turned off at any time, OpenAI said it does not encrypt sensitive information the feature sees, which leaves it vulnerable to hackers.

If Apple pushes forward with AI, it could design more secure versions of features like these in-house, relying on its well-known and trusted commitment to privacy and to protecting user data.

For users who appreciate the subtlety of Siri AI, though, anything similar to ChatGPT or Claude on their devices might not land, unless each feature is opt-in.

Radhika Rajkumar

Senior Editor

Radhika Rajkumar is a senior editor at ZDNET based in New York City. She covers AI, specializing in safety, privacy and security, policy, education, and synthetic media. She also leads ZDNET's newsletter strategy. Radhika holds a Masters in Creative Publishing and Critical Journalism from The New School. See full bio