Google says EU search changes are its biggest quality cut ever
Three stories this morning, and all three are about who gets to decide how a system ranks what you see — Brussels, Canberra, and a boardroom in San Francisco.
Google has started rolling out the EU-ordered redesign of European search results, and a company official told Reuters it marks the largest reduction in service quality in Search's 29-year history. The changes strip real-time prices and other rich features out of the hotel, airline and restaurant carousels, and instead put one specialised search engine — a vertical search service like Booking.com or Expedia — at the top of the page, with two more below it and the stripped-down carousel underneath. Google's argument is that a compliance remedy written to help competitors mostly helps intermediaries: it says earlier DMA changes already drove a 30% drop in free, direct booking traffic to European businesses, and that tests with millions of European users showed high dissatisfaction because people had to retype queries to get what they wanted. Nick Fox, Google's senior vice-president for knowledge and information, said the changes "degrade the user experience for Europeans — boosting online intermediaries at the expense of local businesses."
The context is a €460 million fine in July for favouring Google's own services in shopping, hotels, transport and sports results, followed by a 60-day compliance deadline backed by the threat of periodic penalties up to 5% of worldwide turnover. Two things are worth separating here. The first is that Google is an interested party: every number in this story — the 30% traffic drop, the user dissatisfaction, the "largest in history" framing — comes from Google, and no regulator or independent researcher has validated it yet. The second is that the direction still matters even if the magnitude is exaggerated: this is the first time antitrust enforcement has forced a visible redesign of the world's default interface for finding things, at the exact moment that interface is being rebuilt around AI answers. Google says users outside the EU are unaffected, which quietly makes Europe the largest live experiment in what search looks like when self-preferencing is illegal.
Australia will require social media platforms to let users switch off algorithmic feeds entirely. The Albanese government released draft legislation on Tuesday under the banner "My Feed, My Way," which would force platforms to offer users over 16 a choice between a personalised, algorithm-ranked feed and a chronological feed showing only accounts they follow, plus notifications prompting new and existing users to pick a default. Non-compliance carries penalties of up to A$109.2 million, roughly $79 million. The same bill extends to AI chatbots, online games and other digital services, requiring them to protect under-18s from addictive design features — a notable widening, since the chatbot clause is something no comparable law has yet attempted to enforce.
The timing is awkward for the government, and that is the interesting part. Australia's under-16 social media ban was sold as world-leading, but the eSafety Commissioner's own evaluation found that three months in, 81% of Australian children were still on at least one age-restricted platform — barely below the 86% before the ban — with 58% still using social media daily, against roughly 60% beforehand. This bill is a tacit admission that a ban on accounts did not remove the thing regulators actually object to, which is the ranking system. It also lines up with where the United States landed: Meta's $18 billion settlement with a coalition of states led by California commits it to offering under-18 users a non-algorithmic feed, so the same remedy is now arriving through litigation in one country and legislation in another. Whether anyone chooses the chronological option once prompted is the open question, and the honest bet is that most won't.
Anthropic has walked away from a roughly $6 billion acquisition of Decart after completing due diligence, according to people familiar with the talks cited by Bloomberg. Decart builds software to squeeze more training and inference out of the same chips, alongside world-model work, and the deal would have been Anthropic's largest known acquisition. The distinction in the reporting matters: this was explored and then declined, not agreed and then broken, and Decart stays independent with no confirmed valuation behind the $6 billion figure.
The signal is restraint, and it reads against the rest of Anthropic's year. This is a company that has been committing to compute at a scale that reshapes its balance sheet — we tracked the buildout in Anthropic has signed for 14.8 GW of compute — and maybe $517 billion — so walking away from a $6 billion efficiency-software purchase isn't a cash constraint, it's a judgment that the asset wasn't worth the price after a look inside. That is what due diligence is supposed to do and rarely gets credit for. The likelier reading is that Anthropic would rather own the efficiency work than buy it, or that what diligence found didn't justify the multiple. Either way, the frontier labs are now large enough that their abandoned deals are news, which is its own kind of milestone.
What to watch: whether the European Commission accepts Google's redesign as compliance, or opens the penalty clock.
If a regulator can force a reranking of search, should it be able to force a reranking of an AI assistant's answers? Tell us in the comments.
Sources: Reuters — Google warns of lower quality as it revamps Europe search results · RTÉ — Google warns of lower quality amid Europe search revamp · Techmeme · CNBC — Australia to let social media users 'opt out' of algorithm-based feeds · Prime Minister of Australia — My Feed, My Way · Bloomberg — Anthropic Said to Walk Away From $6 Billion Decart Purchase · Globes — Anthropic decides against $6b Decart acquisition · Tech in Asia — Anthropic drops pursuit of $6b Decart AI deal
