![[Freelance]](../gif/fl3H.png)
A step toward reining in Google
THE UK Competition and Markets Authority (CMA) has required Google to change the way it deals with the “AI summaries” that it is increasingly pushing instead of links to actual journalistic material.
1: A photorealistic depiction of an AI being open about where it gets information from with the authors of that information looking on sceptically, according to ChatGPT, which re-titled the image “robot presenter at a workshop session”
The CMA has designated the search, advertising and “AI” giant as having “strategic market status” in view of its near-monopoly of internet search services in the UK. That allowed it to require, on 3 June, that Google offer publishers “effective tools to prevent their content being used to power AI features in search, such as AI Overviews”. This, the announcement notes, “will put publishers, like news organisations, in a stronger position to negotiate content deals with Google”. The government calls it a “world first”.
Google is also now required “to make sure that publisher content is properly attributed, using clear links, in AI-generated search results”; to explain clearly how it uses and attributes this material; and to “provide publishers with clear and detailed metrics on user engagement with their search content in search generative AI features”. It will have nine months to implement all required changes.
2: A photorealistic depiction of an AI reluctantly being open about where it gets information from with the authors of that information looking on sceptically, according to ChatGPT. The authors named are “Yuval Noah Harari, Noam Chomsky, Jane Goodall, Steven Pinker, Yuval Levin and Faniel Kahneman”. An exposition of the many ways in which this list is interesting will not fit in this margin
Until now, the only way that publishers can remove content from AI-written answers has meant also removing it from actual search results - which as Dominic Ponsford notes in Press Gazette, are the way most people in the UK access the internet.
But it seems to the Freelance that the form of the requirement to allow publishers to exclude content is a win for Google. All it has to do is tweak the specifications for the robots.txt file that provides some crude control over access to a website. Google has fought long and hard to prevent any other means of control being applied.
3: A photorealistic depiction of an AI being forced to reveal where it gets information from with the authors of that information looking on sceptically, according to ChatGPT, which re-titled the image “robot interrogation in tense tribunal”
That idea of negotiting compensation is not going well. New York Times publisher AG Sulzberger told the WAN-IFRA World News Media Congress in Marseille on 1 June that though private investors in the US had sunk $350 billion into “AI” in 2025, “given the small size of deals that have been reported, it appears that less than half of 1 per cent of that investment is going to compensate the people and companies creating the data that powers AI”.
And, in what looks like a pre-emptive response to the attribution requirement, Google on 27 May announced “New ways to find your favorite sources and original content in AI Search” [sic]. You tell it what publications you particularly want to see - and, now, to see actual links to from “AI” summaries. Entirely coincidentally, you thus provide Google with important information about yourself to help it fire ads at you. And it only affects the sites of publications that you already know.
Google is a master of this kind of “sarcastic compliance”. It thereby gains months or years in which it can do what it damn well pleases while regulators such as the CMA respond.
Google is liable as publisher of 'AI' output says German court
UK regulator tells Google to let publishers opt out of AI Overviews pressgazette.co.uk
New York Times chief: How and why publishers should fight AI ‘tsunami’ pressgazette.co.uk
'Artificial Intelligence' our coverage to date