UK CMA’s proposed AI Google Search rules: Explained

On 28 January 2026, the UK Competition and Markets Authority (CMA) proposed a package of measures aimed at improving Google Search services in the UK. If adopted, they could change how businesses win visibility in organic results, how AI answers cite sources, and how easy it is for users to switch search providers.

Key point: These are proposals. The consultation closes at 5pm on 25 February 2026. The CMA will review feedback before deciding what happens next.

What is being proposed

The CMA is consulting on four areas:

• Publisher controls for AI features: More control and transparency over how publisher content is used in AI Overviews, including opt-out options for certain AI uses.
• Fair ranking and transparency: Expectations that ranking and presentation are fair, including within AI experiences, plus a clearer process to raise and investigate issues.
• Choice screens: Steps to make switching search providers easier, including choice screens on Android and in Chrome.
• Data portability: Making it easier for people and businesses to make use of relevant search data.

For businesses, this touches three levers at once: organic visibility, paid search dynamics, and how AI answers influence customer journeys.

Organic search and AI answers: what could change

AI answers reduce the number of clicks for some queries because users get a summary on the results page. If attribution and publisher controls become clearer, being cited and named may matter more for brand demand, even when clicks are lower.

That means “ranking” is no longer the only outcome to track. You also need to know:

• Whether your brand is mentioned in AI answers for your core queries
• Which pages and domains AI answers cite in your category
• Whether competitors are being recommended even when you rank well

If you do nothing else, capture this baseline now so you can measure any impact later.

Fair ranking and issue handling: why it matters for growth

Fair ranking proposals are not about publishing the algorithm. They are about accountability, clearer standards, and a process for businesses to raise concerns when they believe results are being presented unfairly.

How this could affect UK businesses:

• SEO-led teams: If outcomes become easier to challenge, you may see less “unexplained” loss of visibility in some areas over time.
• Paid-heavy teams: If organic becomes more competitive in certain queries, your blended cost per acquisition assumptions may shift. That can be good or bad depending on your mix.
• Comparison-heavy markets: Visibility changes in top-of-page modules and AI answers become more important to monitor, because these areas can influence choice early.

Practical takeaway: treat sudden visibility drops as incidents worth documenting properly, not as normal background noise.

Choice screens: how defaults can affect demand

Defaults drive behaviour. If Android setup and Chrome make it easier to pick a non-Google search provider, even small changes can gradually increase the share of traffic from other engines.

What to do now:

• Make sure reporting splits performance by search engine (traffic, leads, pipeline)
• Build a shortlist of non-Google engines worth tracking for your category
• Create content that performs well in answer-style results (definitions, steps, comparisons)

You do not need to overreact. You do need to be able to see changes early.

Data portability: the long-game measure to watch

Data portability is less visible day to day, but it can lower switching costs and support new search tools, assistants, and competitors.

How this could affect your business over time:

• More places to win intent: You may have more viable surfaces to capture high-intent demand.
• More variability in citations: Different engines and assistants may cite different sources for the same questions.

This increases the value of building content that is easy to interpret and cite, not just content that ranks in one engine.

What UK marketing teams should do next

You do not need a policy strategy. You need a measurement plan and a content plan that works in both classic search and AI answers.

1) Build an AI visibility baseline

• List 30 to 50 buyer questions across your funnel
• Record whether AI answers appear, who is cited, and whether your brand is mentioned
• Repeat for your top competitors and for “category leaders” that get cited often

2) Make key pages easier to cite

• Put a plain-English definition near the top of core pages
• Answer the question directly, then expand with detail and examples
• Use bullet lists for steps and decision criteria
• Add evidence where possible (numbers, methods, worked examples, customer outcomes)

3) Tighten technical clarity

• Fix indexing and canonical issues that create duplication or confusion
• Use structured data where it genuinely fits your content
• Improve internal linking so your strongest pages are easy to find

4) Update your reporting

• Keep: rankings, clicks, conversions
• Add: share of mentions in AI answers, share of citations, and which domains get referenced
• Connect to: pipeline and revenue influence, not just sessions

Where Clavius fits: Clavius by Tilio helps you track brand mentions and competitor mentions in AI-generated answers, spot which sources get cited in your category, and prioritise the content and technical fixes most likely to improve visibility.

• Confirm internal owners for SEO, paid search, analytics and content
• Create a 30 to 50 query baseline for AI answers (mentions, citations, sources)
• Identify your “citation competitors” (often different from product competitors)
• Refresh core pages for clarity: definitions, steps, comparisons, evidence
• Run a technical audit: indexing, canonicals, structured data, internal linking
• Split reporting: rankings and traffic versus AI mentions and citations
• Track changes monthly so you can see trend movement, not one-off noise

FAQs

Is this only relevant to publishers? No. If AI answers cite sources more explicitly, any business that uses content for demand generation can benefit from being named and cited.

Will this change traffic overnight? Unlikely. If changes happen, they usually roll out in phases. That is why baselining now matters.

Should we change our SEO strategy right now? Do not panic-change. Improve measurement first, then make pages clearer and easier to cite.

What is the most useful action this week? Build a baseline for your top buyer questions: AI answer presence, citations, and whether your brand is mentioned.

How do we know if we are winning in AI answers? Track share of mentions and citations for your category queries over time, and correlate that with brand search growth and pipeline.