Apple backs Google against EU AI access rules

Apple has quietly sided with Google in a public pushback against draft European Union rules that would force platform owners to grant third‑party AI services deeper access to on‑device and assistant features. The alignment was reported in filings and coverage on May 13–14, 2026.

Brussels opened a specification process under the Digital Markets Act (DMA) in late January and issued draft measures in April aimed at ensuring third parties can interoperate with key Android capabilities. The Commission invited feedback and set a timetable for a final specification decision later this summer.

In written submissions, Apple warned the draft measures “raise urgent and serious concerns,” saying they could create “profound risks” for user privacy, security, safety, device integrity and performance if implemented as drafted. That language and Apple’s position were reported after the Commission’s consultation window closed.

Google has also pushed back, arguing the proposals risk undermining security and create technical burdens for platform maintainers while describing some measures as an unwarranted intervention. The company says the changes could destabilize protections built into the operating system. Those objections form the public core of the platform response to Brussels.

Apple’s intervention matters because the company is itself subject to DMA-driven interoperability demands in Europe and is already rolling limited compliance changes for iPhones and iPads. That legal backdrop helps explain why Apple framed the Google‑focused consultation as one with broad implications for multiple ecosystems.

At stake are technical hooks and data flows that let assistant services do things such as read and act on messages, access calendars, trigger device sensors, or integrate with mapping and payment apps. The Commission’s draft measures describe ways third‑party agents might be allowed equivalent access to such capabilities to ensure contestability.

Platform owners paint that capability as a security and safety risk. Their submissions say opening low‑level APIs and system services to external agents increases attack surface and could allow malicious actors to exploit sensitive device functions. Those claims are central to the industry push for “guardrails” or narrower remedies.

Regulators and many EU governments offer the opposite argument: requiring access and interoperability will prevent platform owners from favouring their own AI services and will expand consumer choice across devices and assistants. Brussels views contestability as a remedy for concentrated power in AI‑enabled platforms. The Commission says the proposed measures aim to give Europeans more choice in on‑device AI.

The wider political context sharpens the dispute. Several EU capitals have also urged Brussels to beef up technological sovereignty and protections against high‑capability models developed primarily outside Europe, citing risks to critical infrastructure and democratic resilience. Those concerns are part of why lawmakers are moving aggressively on platform rules.

From a developer and startup angle, the debate is twofold: some third‑party AI providers want guaranteed technical access to build competitive agents, while others warn that rushed or poorly specified interoperability can entrench dominant services or create compliance costs that favor incumbents. The result is a complex tradeoff between opening ecosystems and preserving secure, predictable platforms.

What happens next is procedural but consequential. The Commission will review the feedback it received, may revise the draft measures, and must adopt a final set of specifications under the DMA timetable by late July 2026. Any final decision will likely trigger legal and political responses on both sides of the Atlantic. The clash over the specifics of model access, developer APIs and device sovereignty is now a front in an unfolding transatlantic policy fight.