The European Union has spent the past few years pushing technology companies to open up different parts of their ecosystem, from charging ports and app stores to messaging platforms. Now, it is turning its attention to something newer, and arguably more central to how smartphones will work in the future: artificial intelligence.
In its latest move, the European Commission has told Google that Android needs to offer competing AI services the same level of access as its own Gemini assistant.
What the EU is actually asking for
The EU’s position is relatively straightforward. It believes Google is giving Gemini an advantage that goes beyond being pre-installed, and that this advantage is now shaping how AI works on Android devices.
Today, Gemini is deeply integrated into the operating system. It can interact with apps, access on-device context, and perform tasks that extend beyond a single interface. Third-party AI tools, even when installed, operate within far narrower limits.
The Commission’s proposal is to change that balance. Its preliminary findings outline a framework where competing AI services should be able to interact with apps, access device-level context, and execute tasks across the system in a similar way.
Why this matters
For most of the smartphone era, apps have been the primary way users interact with their devices. Operating systems controlled access, while apps competed within those boundaries. That model is beginning to shift. AI assistants are increasingly positioned as the layer that sits above apps, capable of moving across them to complete tasks.
With AI systems such as Google Gemini offering agentic capabilities, users can ask the assistant to act on their behalf, whether that involves sending a message, sharing content, or retrieving information. This removes the need to open each app individually to take an action.
This changes where control sits within the ecosystem. If one assistant has deeper access to system features and data, it becomes the default interface through which the device is experienced. The EU regulators argue that over time, this advantage can become structural.
Google’s argument
Google has pushed back against the proposal, calling it “unwarranted intervention” and warning that it could undermine how Android is designed to work.
The company’s position centres on the idea that device makers should retain control over how AI services are implemented. It has also raised concerns that opening system-level access could expose sensitive hardware and user data, increase costs, and create new security risks.
“This unwarranted intervention would strip away that autonomy, mandate access to sensitive hardware and device permissions; unnecessarily driving up costs while undermining critical privacy and security protections for European users,” said Google senior competition counsel Claire Kelly, according to a report by Ars Technica.
What happens next
At this stage, the EU’s proposal is not a ruling. The commission has opened the measures to public consultation, inviting feedback from industry participants, developers, and other stakeholders. Interested parties have until May 13 to submit their views on the proposed measures.
The Commission has said it will assess responses from both stakeholders and Google before deciding whether any adjustments are needed. It also said that a final decision must be adopted within six months from the opening of the start of the proceedings, and would outline binding measures that Google would be required to implement.
Part of a broader EU playbook
The current move is not a standalone intervention. It sits within a broader regulatory framework under the Digital Markets Act, which was first adopted in 2022 and came into full effect in 2024.
At its core, the DMA targets big tech companies designated as “gatekeepers” — platforms that act as critical intermediaries between businesses and users. When a platform controls access to distribution, data, or core functionality, it can shape outcomes across an entire ecosystem.
Most of the changes enforced under the DMA so far have focused on reducing that control at different layers of the stack.
On the software side, companies like Google and Apple have been required to allow alternative billing systems in app stores, introduce choice screens for search engines, and limit how data is shared across services. This goes beyond operating systems, as companies such as Meta have also been asked to enable interoperability between messaging services, including opening up WhatsApp to interact with other platforms.
On the hardware and device side, broader EU regulations, working alongside the DMA, have pushed for standardisation and user control. This includes mandating USB Type-C charging, improving repairability standards, and extending expectations around software updates and device longevity.
Taken together, these measures point to a consistent strategy. Rather than breaking up companies or imposing blanket restrictions, the EU is focusing on specific control points — charging ports, app stores, messaging networks, and now AI.
What changes if the EU gets its way
If implemented, the proposed measures could significantly alter how AI functions on Android devices. Third-party assistants would no longer be limited to their own apps, but could begin to operate across the system, interacting with other apps and responding to user context in real time.
This would bring them closer to the capabilities currently associated with Gemini, effectively reducing the gap between Google’s own assistant and competing services. It could also encourage a broader range of AI experiences, as developers gain the ability to build deeper integrations rather than standalone tools.
In the meantime, device makers are already moving toward more layered AI experiences. For example, Motorola has integrated Perplexity AI and Microsoft Copilot directly into its Moto AI ecosystem on new devices. Likewise, Samsung has been integrating its own Bixby assistant across the software level, while offering third-party services such as Perplexity pre-installed on its new flagship devices.
Even so, Gemini remains the assistant with the deepest level of system access. It is the one capable of interacting across apps, understanding device context, and delivering what is increasingly described as agentic functionality.
Other AI tools may coexist on these devices, but they operate within narrower boundaries. They can respond, but they cannot act in the same way.
If the EU’s approach begins to reshape how Android handles AI access, that distinction could start to change.
Will these changes also affect Indian users
For consumers, particularly in markets like India, the immediate impact of such changes may not be obvious. But over time, shifts driven by European regulation have a way of extending beyond the region.
This has already played out across multiple layers of the smartphone ecosystem. When the EU mandated a common charging standard, Apple moved to USB Type-C with the iPhone 15 series. While the requirement was specific to Europe, the change was implemented globally. Similarly, rules around repairability and battery access have pushed manufacturers to improve device design standards across markets, not just within the EU.
The reason is largely practical. Hardware changes are difficult to regionalise. Building separate versions of devices for different markets increases cost and complexity, which makes global alignment the easier route.
Software, however, tends to behave differently.
In cases involving app stores and digital ecosystems, regulatory changes have not always translated directly across regions, but they have influenced parallel action elsewhere. In India, the Competition Commission of India has already taken on both Google and Apple over similar concerns around platform control. The Google Play Store case resulted in penalties and mandated changes around billing and app distribution, while Apple’s App Store practices are currently under scrutiny over in-app payments and ecosystem restrictions.
These cases are not identical to what the EU is proposing around AI, but they reflect a similar line of thinking — that control at the platform level can shape how entire ecosystems function.
For now, there is no indication that India is looking to regulate AI on smartphones in the same way. But the pattern is familiar. Regulatory moves in one region often set the direction for conversations in others, especially when they involve companies operating at a global scale.
