This guest article by Stuart Forrest was first published on Flashes & Flames: The Global Media Business Weekly.

Has the UK gone soft on Google?

The UK’s Competition & Markets Authority (CMA) has unveiled what it terms “world first” measures, purportedly granting publishers new-found agency over the utilisation of their content within Google’s AI Overviews and AI search functionalities.

The UK-only legislation is framed by the CMA as “a fairer way” for publishers to negotiate deals with Google. Simultaneously, Google launched an analytics trial for UK publishers to assess AI search visibility via Google Search Console. Google, promised future global access to these granular insights and controls, though without a firm timeline.

While action here is certainly preferable to stagnation, there is a lingering concern that these measures provide Google with a soft landing, potentially establishing a dangerously lenient regulatory bar just as other global jurisdictions weigh the requirements for a more meaningful future value exchange between publishers and the platforms.

After CMA’s decision to designate Google “strategic market status” (SMS) in the UK, allowing them to introduce targeted rules over how they operate, the recent announcement unveiled the first controls under this new regulatory status. The key announced elements were:

• Tools to allow publishers to exclude their content from AI features in search results pages at domain or URL level. Until now, publishers haven’t been able to opt out of AI features without losing all search traffic.

• A requirement by Google to properly attribute content in AI-generated results, with clear links.

• The ability to opt-out of content use in “fine tuning” of models, where foundation models are built upon with additional training data.

Separate from the regulatory controls but complementary to them, Google subsequently announced a limited trial for UK site owners of new AI search feature performance data within Google’s Search console analytics tool, for the first time giving insights into how AI features compare to regular search links.

All these changes are for UK search results only, and Google has six months to implement the controls. Some publishers are already seeing the data insights in their Search Console accounts and CMA’s Will Hayter has suggested that AI feature control tools may also arrive sooner than the December 2026 deadline.

But is it a victory for publishers, or a soft landing for Google?

The news was greeted with cautious optimism by large publishers and their trade associations. But is this actually the good news CMA believes, or is it a regulatory compromise that provides a Round One victory for Google by setting the regulatory bar too low?

The intricate details reveal major compromises that may prevent real change. This situation risks benefiting Google, which wants to prove AI features help both owners and users. If these hard-won controls see low adoption, it bolsters Google’s case for minimal regulation.

There are two critical reasons why these controls may fail to see any kind of meaningful uptake:

1. Publishers must opt-out of AI features, not opt-in

While UK publishers lobbied for an opt-in model, the CMA has instead mandated an opt-out framework for these AI functionalities. Consequently, the responsibility shifts away from Google to prove the worth of its AI features and onto publishers and trade groups, which must now evaluate the benefits of staying included and ensure that these new options are widely understood within the industry.

Given the intense competition and declining traffic volumes facing many in the sector, there is concern that a combination of market pressure and operational inertia may prevent many publishers from actually utilizing these newly available controls.

The danger is that low adoption of the controls, despite heavy lobbying, will allow Google to argue to regulators that its search features – which many blame for falling referral traffic – benefit both users and publishers.

2. Newly provided AI search data insights are simply insufficient

Publishers wanting to assess AI’s impact on referral traffic will find Google’s new data insufficient for informed decisions. While a slight improvement, the data lacks click metrics, query intent details, and pre-May 2026 historical records.

Without click data or comparative history, publishers simply cannot judge the value of inclusion or link AI rollouts to traffic declines, leaving Google’s transparency efforts largely illusory.

It’s notable that Bing – while a minnow in search terms – has actually provided substantially more insight to publishers for some months; historical data and aggregated search query data at URL level.

What does this mean for US publishers?

Google has stated that, while the proposed controls and greater data transparency will be tested in the UK first, they will eventually roll-out globally, albeit with no specific timeline commitment.

This is a good thing for US-based publishers, where there’s no meaningful legislative agenda favouring greater controls for publishers and no equivalent legislative instrument to the UK’s “SMS” designation for Google allowing such controls to be enforced.

In the US, the conversation is at individual publisher level, with the industry bifurcated between those large enough to pursue action against Google and AI platforms (eg New York Times and Condé Nast) or take up the small number of paid content deals on offer, leaving the vast majority of small publishers without either meaningful protection or licensing deals.

Will the UK actions have an impact?

Possibly, but certainly indirectly. If large numbers of UK publishers do withdraw their content from AIO then Google’s search experience may well suffer, and this could force a more equitable content licensing conversation with publishers globally, as those controls are made available to them and content is withdrawn in other jurisdictions.

Which brings us back to the critical question; has the CMA done enough to allow UK publishers to make a decision to exercise these controls?

The short answer really is no.

The opt-out measure, compounded by insufficient data granularity, greatly favours the status quo and it feels unlikely that a critical mass of publishers will withdraw their content. That’s a crucial requirement to bring platforms to the table to discuss meaningful licensing models.

Whatever happens, publishers and their associations will be watching closely – and should not concede anything to Google at what might prove to be a critical moment in this industry’s attempts to secure a meaningful future value exchange with platforms. Time to hold your nerve.

Good reads: