The most consequential thing about Apple's WWDC 2026 announcements on 8 June is not a feature — it is a map. The new Siri AI and the next generation of Apple Intelligence will not arrive the same way everywhere. Siri AI is delayed on iPhone and iPad in the European Union, unavailable in China pending regulatory requirements, and launching alongside a new age-based child-safety layer that reflects tightening online-safety regulation. For anyone tracking technology governance, the keynote made clear that what Apple can ship is now decided as much by regulators as by engineers.

The EU: Siri AI delayed on iPhone and iPad

In a same-day notice, Apple said that because of the Digital Markets Act it would not ship Siri AI in the EU with the release of iOS 27 and iPadOS 27. EU users will still get Siri AI on macOS 27, visionOS 27 and watchOS 27, and developers based in the EU will not be able to build or test Siri AI features on iPhone and iPad. Apple gave no timeline for when, or whether, the iOS and iPadOS versions will arrive.

Apple's framing is pointed. It argues that, under what it calls EU regulators' extreme interpretation of the DMA, it would have to give rival virtual assistants access to the same private-data and app-control capabilities that Siri AI uses, without safeguards Apple considers adequate — a risk it says security researchers have shown can be abused to steal passwords and photos. Apple said it proposed an intermediary it calls a Trusted System Agent and an 18-month phased rollout, and that the European Commission rejected every proposal. Craig Federighi said Apple was "deeply disappointed that our EU users won't have Siri AI on iPhone or iPad," and that the company has no current timeline for the EU.

The other side of the DMA fight

That is Apple's account, and a governance reader should weigh the regulator's rationale against it. The DMA's interoperability requirement — Article 6(7) — obliges designated gatekeepers to give third parties access to the same operating-system features Apple uses for its own services, so that competitors can, in the Commission's words, compete on equal terms. The Commission's position is that this is the price of being a gatekeeper, not an attack on privacy.

Critics of Apple's framing go further, arguing that a gatekeeper cannot use privacy as a blanket reason to block competition, and noting that Apple has repeatedly warned of dire security consequences from the DMA yet has still shipped the required changes, from alternative app stores to NFC access, without the predicted harm. The Free Software Foundation Europe, reviewing Apple's request-based interoperability process, reported that as of 22 March 2026 none of the 56 formal developer requests Apple had received under Article 6(7) since May 2025 had produced a new interoperability solution; the European Commission had opened a specification case to force a fairer process. The honest read is that both sides hold something real: the attack surface Apple describes is a genuine risk as assistants gain the power to act across apps, and the entrenchment the EU describes is a genuine harm when one firm controls the platform. This is a values conflict, not a simple case of obstruction by either party.

China: approval as the gate

China is the more direct case. Apple stated plainly that Siri AI and the other new Apple Intelligence features will not be available there at all while it works through regulatory requirements. Where the EU dispute is about the terms of access, China's is about approval itself: the country requires generative-AI services to clear regulatory filing and approval before launch, which routinely gates foreign AI features and typically forces reliance on locally approved arrangements. The result is that Apple's most prominent 2026 feature is simply absent in one of its largest markets, with availability dependent on a regulatory process Apple does not control.

The child-safety layer

Running underneath the AI news is a second governance theme. Apple previewed a set of child-safety features that make a child account required for children under 13 and available up to age 18, enabling age-tailored safeguards across the system. Its Communication Safety feature, already on by default for under-18s to blur nudity, will now also intervene on shared images or videos containing gore or violence, and a new Declared Age Range API lets developers tailor app experiences to a child's age band without the child sharing a birthday.

These features land into a global wave of child-safety and age-verification regulation, and they carry their own governance tension. Age-based protection requires some form of age signal, which sits awkwardly against data-minimisation principles; Apple's answer is to frame its age-range approach as privacy-protective by sharing a range rather than a birthdate. Whether that satisfies regulators pushing for stronger verification, and privacy advocates wary of more age data, is an open question — but it is another instance of regulation shaping the product surface directly.

What this means for the region

For Southeast Asia, the practical lesson is that AI availability now tracks each jurisdiction's regulatory posture. An English-language market such as Singapore is positioned to receive Siri AI when the EU will not, simply because its regulatory environment does not impose the DMA's interoperability mandate; Vietnamese is the only Southeast Asian language on Apple Intelligence's initial supported-language list. The wider signal for regional regulators is double-edged: a lighter-touch stance tends to get features sooner, but the EU's experience shows that a large enough market can force a global company to negotiate terms it would not otherwise offer.

Key Takeaways

  • Apple will not ship Siri AI in the EU on iOS 27 and iPadOS 27 (it will on macOS, visionOS and watchOS), citing the DMA; EU developers can't build Siri AI features for iPhone or iPad, and Apple gave no timeline.

  • Apple blames what it calls EU regulators' extreme interpretation of the DMA and says its Trusted System Agent proposal and 18-month rollout were rejected; the Commission's basis is Article 6(7), which requires gatekeepers to open the same features to rivals.

  • In China, all the new Apple Intelligence and Siri AI features are unavailable while Apple works through the country's generative-AI approval regime.

  • A new child-safety layer (child accounts required under 13, Communication Safety extended to gore/violence, a privacy-oriented Declared Age Range API) reflects rising age-verification regulation worldwide.

  • For ASEAN, AI availability now tracks regulatory posture: English-language Singapore is set to get Siri AI where the EU won't; Vietnamese is the lone Southeast Asian language in Apple Intelligence's initial list.