EU’s New Rules Shake Up Google and Apple’s AI Playbooks

Google and Apple have long had a comfortable lead in the artificial‑intelligence arms race, simply because they own the hardware that almost everyone carries in their pockets. Roughly five billion Android phones and iPhones are active worldwide, a figure quoted by market‑research firm Omdia, which gives the two giants a built‑in distribution channel that most start‑ups can only dream of.

That monopoly, however, is now under the EU’s microscope. The European Commission, acting under the Digital Markets Act (DMA) – a law that finally stopped being a vague idea in 2023 – wants to make the smartphone AI market less of a closed club. By July 2027, Google must let any third‑party AI assistant talk to Android the same way its own Gemini does, and Apple will have to let rivals tap into Siri’s microphone, location and app‑control abilities.

What does that look like in practice? Imagine you pick up your phone and, instead of saying “Hey Google,” you could say “Hey ChatGPT” and have the AI not only answer a question but also order an Uber, book a table, or turn on your smart lights – all without Google’s code in the background. The EU says this will give smaller AI firms a fighting chance and, frankly, give users more choices.

Google isn’t thrilled. In a statement, Kent Walker, president of Global Affairs at Alphabet, warned that handing “deep permissions” to outside services could open security holes, dilute user privacy, and even threaten national‑security secrets. Apple echoes the sentiment, pointing out that the DMA would force Siri to expose private data to competitors – a move the company calls “unacceptable.” The tech giant has already delayed its next‑generation Siri rollout in Europe, citing the new regulations as the stumbling block.

Security experts agree there are genuine risks. Calli Schroeder of the Electronic Privacy Information Center cautions that an AI with unrestricted access to messages, microphones, and screens could silently harvest data if safeguards aren’t airtight. Michael Stokes, senior VP at Veilant, adds that a misbehaving assistant could cause real‑world harm, from sending messages to the wrong contact to revealing your location.

On the flip side, the rules also oblige Google to share portions of its search data with rival search engines and AI chatbots. That data, once a closely‑guarded secret, could help newer players like OpenAI fine‑tune their models for European users. Omdia analyst Runar Bjorhovde notes that this could slowly erode Google’s stranglehold on the mobile AI ecosystem.

Manufacturers such as Samsung will be the gatekeepers, deciding which AI assistants get the coveted permissions and policing their behavior, according to Google. Whether that model works in practice remains to be seen; the EU is betting that competition will force better privacy‑by‑design solutions, while critics argue the tech giants have historically put profit ahead of privacy.

In short, the DMA is set to rewrite the rules of the AI game on smartphones. It may give ChatGPT a larger foothold in Europe – it already sits on about 30 % of devices – and force the big two to rethink how tightly they cling to their data troves. The outcome will likely shape how much of our personal lives AI can safely manage in the years ahead.