Europe may be loosening Google’s grip on mobile search and AI, but it is not bringing back the old internet. Telemedia must prepare for services discovered by machines, bought through multiple payment rails and consumed without a conventional website visit, says Paul Skeldon
The European Commission’s decision to force Google to open parts of Android and share anonymised search data with rival search engines and AI chatbots looks, at first glance, like good news for anyone who has become too dependent on Google for traffic.
And it is – but not quite in the way the telemedia market might hope.
The ruling could create more search services, capable assistants and discovery routes. It will not restore cheap clicks from blue links. Instead, several AI agents may compete to recommend, compare and eventually operate services.
For telemedia, this lands just as two other changes are undermining its traditional acquisition model: Google Zero and HTTPS-by-default.
The old funnel is being squeezed at both ends
The classic mobile-content journey was beautifully simple: a Google ad or search result led to a landing page to a network identification and ended in a two-click DCB subscription.
Google Zero – where no one clicks beyond the Google/Gemini result – weakens the first part. In the first four months of 2026, 68% of US Google searches ended without a click, according to SparkToro’s analysis of Similarweb data. Separately, Pew Research found that users clicked a conventional result in only 8% of visits containing an AI summary, against 15% when no summary appeared. Google is still answering the question, but increasingly it is keeping the consumer.
HTTPS-by-default attacks the other end of the funnel. Chrome is moving towards secure connections as the default, with the setting already enabled for Enhanced Safe Browsing users and due to become standard in Chrome 154 in October. That is good for security, but encrypted traffic prevents the network-level header enrichment on which many frictionless DCB journeys have depended.
So, Google Zero changes how the customer arrives, while HTTPS-by-default changes whether that customer can be identified and charged once they get there.
Fewer visitors multiplied by more checkout friction can produce a disproportionately severe fall in subscriber acquisition. The old model relied on Google sending the user to the merchant and the operator silently recognising that user. Both assumptions are weakening at once.
Europe is opening the market, not restoring the click
This is why the EU ruling matters. Qualifying rivals will gain access to anonymised query, ranking, click and view data that can help them improve search and retrieval. Third-party assistants on Android will eventually be able to respond to voice activation, understand on-screen context and perform actions inside apps.
That may prevent Gemini from becoming the only meaningful AI gateway on Android and give telemedia companies new distribution partners.
However, none of those assistants has a commercial incentive to send unnecessary traffic to a merchant site. If an AI can explain a service, compare alternatives, check compatibility and initiate the transaction, the website becomes a data source or fulfilment endpoint rather than the centre of the journey.
The new customer path may look more like: AI request leads to a service recommendation that garners an eligibility check, which demands explicit consent, the selection of a payment rail and then activation. The consumer may never see a conventional landing page.
Telemedia marketers must now ask a different question: what would make an AI confidently recommend and execute this service?
That requires machine-readable information covering the proposition, price, renewal, trial terms, cancellation, supported networks, age suitability, payments and support. It must be consistent across merchant sites, MNO portals, app stores, affiliates, reviews and social channels.
Ambiguity becomes commercially dangerous. A vague price, unclear recurring charge or difficult cancellation process is no longer merely a compliance problem after acquisition. It can become a machine-readable reason not to recommend the product at all.
Reputable providers may benefit. Recognisable ownership, licensed content, transparent terms, good reviews and clear support give an assistant evidence on which to base a recommendation. Trust becomes part of discoverability.
AI can reproduce a horoscope or answer trivia without sending anyone elsewhere. It cannot so easily replace licensed sport, live participation, interactive gaming, specialist expertise, communities or network services. The dividing line is between information that can be answered and experiences that must be accessed.
Traffic may shrink, but intent could improve
There is an upside. Anyone reaching a service after an AI-led conversation may already understand its price, renewal, content, network requirements and payment choices. That could mean fewer impulsive subscriptions but higher intent, better retention and fewer cancellations, refunds and complaints.
But better traffic does not repair a broken payment journey. DCB must evolve from a browser-dependent mechanism into a secure, API-accessible payment rail. CAMARA and GSMA Open Gateway-style number verification, authentication and carrier-billing APIs could allow a merchant, app, messaging platform or AI assistant to check eligibility, obtain consent and initiate billing without an unencrypted HTTP page.
DCB must also sit inside a multi-rail offer. An agent may choose carrier billing for one customer, a stored wallet for another and account-to-account payment for a third. If DCB is not equally callable, it risks being designed out of the new interface.
Engagement must move beyond borrowed traffic
Telemedia providers must avoid swapping dependence on Google for dependence on a dozen AI assistants. They need stronger direct routes through MNO apps and portals, RCS, SMS, WhatsApp, push notifications, email, creators, app stores, device partnerships and bundles.
Affiliates with recognised audiences and communities will be more defensible than those built on SEO arbitrage. MNOs become more valuable because they combine trust, communication, eligibility data and billing. This is a key point and a key pointer to the future not just of telemedia services, but of telcos too. As we discussed last time around providing user ID and protecting consumers from harm, telcos have an opportunity to play a much more pivotal role in the discovery and traffic generation process.
Measurement must change too. Click-through and last-click conversion reveal less when discovery happens inside AI and purchase occurs later elsewhere. AI visibility, assisted conversion, activation cost, retention, lifetime value and complaints will tell a more useful story.
The EU has not saved the old telemedia funnel. It may have ensured that Google will not be the only company replacing it.
The winners in the next phase will not necessarily be those generating the most traffic. They will be the services that machines can understand, consumers can trust, operators can verify and multiple payment rails can complete. Customers will increasingly find telemedia through an AI recommendation, an operator environment, a creator or a conversational message – and may subscribe without ever “visiting” the service in the traditional sense.
The age of the click is giving way to the age of selection. Telemedia’s task is to make sure it remains selectable.
