Illustration created with Canva AI by Author — a UK flag-handled magnifying glass examines an AI circuit board, symbolising the CMA's new scrutiny of Google Search and AI-generated results under the UK's 2026 digital markets conduct requirements.
The UK is entering a defining moment in digital governance. In June 2026, the Competition and Markets Authority (CMA) imposed new legal requirements on Google Search that reach beyond traditional rankings and into AI Overviews — the AI-generated answers that increasingly sit above the familiar list of links.
This is more than a competition story. It is a sign that AI-powered search is no longer being treated as a product experiment that platforms can adjust in private. It is now being treated as public-facing infrastructure that shapes what people see first, what they trust first, and in many cases, where they seek help first.
For mental-health professionals, service providers, and developers of digital care tools, this matters because search is often the first doorway into support. If AI is summarising information about anxiety, trauma, depression, or therapy options, then the rules that govern those summaries matter just as much as the underlying technology.
SIDEBAR: Key terms in plain language
AI Overviews: Google's AI-generated summaries that can appear above traditional search results. The CMA's June 2026 measures explicitly treat them as part of Google Search rather than as a separate experimental layer.
Fair ranking: The CMA's fair ranking requirement says Google must rank organic search results using objective and non-discriminatory criteria, and that obligation applies to AI Overviews and other generative search features as well.
Publisher opt-out: The CMA's June 2026 publisher measure requires Google to let publishers opt out of use in AI search summaries while remaining in ordinary search results.
Attribution: The same CMA action requires clearer attribution and linking back to original sources in AI-generated search results, which matters because traceability is a basic condition of trust.
Conduct requirement: A binding rule imposed by the CMA on a company with strategic market status, setting out how it must behave in a specific digital market.
What the ruling says — in plain language
At its heart, the CMA's action is about one thing: balance. The regulator is not banning AI search. It is trying to ensure that innovation in search happens with clearer safeguards for businesses, publishers, and the public.
The CMA has imposed two especially important conduct requirements on Google.
1. Fair ranking requirement
Under the fair ranking requirement, Google must rank organic search results using objective and non-discriminatory criteria, and this requirement explicitly includes AI Overviews and other generative AI search features.
Google must also:
- Provide greater transparency about how rankings work.
- Give advance notice of significant ranking changes that could materially affect businesses.
- Create clear processes for businesses to raise concerns about ranking decisions and have them addressed effectively.
In simple terms, this means Google cannot treat major search changes as a complete black box in the UK, especially when those changes affect visibility, traffic, and commercial survival.
2. Publisher controls for AI search
Earlier in June, the CMA also required Google to give publishers stronger control over how their content is used in AI search features. That includes:
- Allowing publishers to opt out of having their content used in AI Overviews and related AI features while still remaining in ordinary search results.
- Allowing publishers to prevent their content from being used to train or fine-tune standalone AI models beyond core search use.
- Requiring clearer attribution and linking back to original sources in AI-generated results.
This is widely seen as a world-first move because it separates ordinary indexing from AI reuse. Until now, many publishers felt they had to accept AI summarisation if they wanted to remain visible in search at all.
CALLOUT: Why mental health is different
Mental-health searches are often made under strain, uncertainty, or isolation, which means the first answer a person sees may shape both belief and behaviour more strongly than in lower-stakes topics. The CMA's decision matters here because it confirms that AI summaries are not outside the search system; they are part of it, and must therefore sit inside rules about fairness and transparency too.
When someone searches "panic attack symptoms", "am I depressed", or "find therapy near me", the search page functions like an informal triage layer. If that layer is increasingly AI-mediated, then source visibility, attribution, and contestability stop being abstract technical issues and become part of the care pathway.
Why this matters for AI in mental health
Mental-health information is not neutral content. It often reaches people when they are distressed, uncertain, isolated, or trying to make sense of symptoms they may not yet have named. In that context, the first explanation a person sees can shape not only what they believe, but what they do next.
Search is now part of the care pathway
For many people, search has become an informal triage layer. Someone types "panic attack symptoms", "therapy for trauma", "is this depression", or "how to find a psychologist near me", and the results page becomes an entry point into decision-making.
As AI Overviews expand, that entry point changes. Instead of comparing several sources, users may encounter a single AI summary first — one that feels authoritative even when it is condensed, incomplete, or unevenly sourced.
The CMA's ruling matters because it says those AI summaries are not outside the search system. They are part of it, and they must therefore be governed by fairness and transparency rules too.
Trust in health information depends on explainability
When AI surfaces mental-health information, trust does not come only from fluency. It comes from traceability: where did this answer come from, why was this source chosen, and can the user reach the full context?
The publisher requirement helps here. By requiring attribution and clearer links in AI-generated results, the CMA is pushing search closer to an accountable information environment rather than a closed answer box.
For mental-health organisations, charities, clinics, and public-health services, this matters because it becomes easier to see whether their material is being surfaced, how it is being represented, and whether people can actually get back to the original source.
Human judgement still matters
Although the ruling is not specific to healthcare, it supports a broader principle that is especially important in mental health: systems that shape important outcomes should not operate beyond meaningful challenge.
If a search update suddenly buries a specialist trauma clinic, a local psychology service, or a crisis-support organisation, there is now meant to be a clearer path to question that change in the UK.
That does not mean every dispute will be solved. But it does mean that AI-driven visibility is beginning to move from untouchable platform behaviour to something answerable to public rules.
CALLOUT: High-stakes queries to watch
These are the kinds of searches where AI summary design matters most:
- "Am I depressed?"
- "Panic attack or heart attack?"
- "Therapy for trauma near me"
- "Postpartum depression help"
- "How to stop self-harm"
- "Domestic violence counselling"
- "Suicidal thoughts help now"
In queries like these, answer quality is not only about factual correctness. It is also about whether the system reliably shows trusted sources, points people toward local or crisis support where relevant, and allows the user to reach the original context rather than remaining trapped in a synthetic summary.
What this means globally
The UK is not the only jurisdiction concerned about platform power, but it is among the first to apply detailed digital markets powers directly to search rankings and AI summaries.
The European Union already has broad gatekeeper obligations under the Digital Markets Act, and US regulators continue to pursue antitrust cases against Google. What makes the UK move distinctive is that it links competition law, ranking fairness, AI summaries, publisher controls, and complaint mechanisms into one operational package.
That makes the UK a test case. If these rules prove workable, other regulators will have a live example of how to regulate AI search without demanding full public disclosure of Google's ranking systems.
For mental-health AI, this matters because it shifts the conversation globally. It becomes easier to argue that health-related AI answers should be fair, attributable, contestable, and open to oversight — not just technically impressive.
SIDEBAR: What is evidenced, what is emerging, what is speculative
Evidenced: The CMA has imposed a publisher-related conduct requirement and a fair-ranking conduct requirement on Google Search in the UK. Both measures expressly reach into AI-driven search features including AI Overviews.
Emerging: What remains to be seen is how these rules will work in day-to-day practice — how clearly Google will explain ranking changes, how useful complaint channels will be for smaller organisations, and how visible attribution will become inside AI answers.
Speculative: It is plausible that some compliance features built for the UK — such as clearer attribution or more granular publisher controls — could spill into other markets over time, but that outcome is not guaranteed by the CMA's orders themselves.
What this means for Australia
Australia has not yet imposed the same type of legal obligations on Google Search. But the conditions that made the UK act are very familiar here.
Australian regulators have long raised concerns about the opacity and power of digital platforms, and the ACCC has repeatedly argued for greater transparency and oversight of Google's operations.
At the same time, Australian businesses, clinics, and service providers are deeply dependent on Google Search, Maps, and platform visibility. Reporting has already described how AI-driven changes to Google Search have affected traffic patterns for businesses that rely on discovery through the web.
For Australian mental-health providers and digital-health founders, the UK ruling matters in three ways:
- It offers a working model of what search transparency rules could look like.
- It gives Australian advocates a concrete precedent when arguing for oversight of AI search.
- It increases the chance that some Google compliance features — such as clearer attribution or better publisher controls — may eventually spill over into other markets.
In that sense, the UK decision does not just affect British publishers. It changes the policy imagination elsewhere.
SIDEBAR: Questions for Australian providers
For Australian clinics, charities, and digital-health founders, the UK ruling is useful because it turns a vague concern about platform opacity into a concrete governance model. Australia has not yet imposed the same legal structure on Google Search, but the ACCC has long documented concerns about digital-platform power, information asymmetry, and weak bargaining positions for dependent businesses.
Questions worth asking now:
- How much of our discovery depends on Google Search, Maps, and Business Profiles?
- Which of our pages are most vulnerable if AI summaries reduce click-through?
- Are our crisis, intake, and service pages clearly attributable and easy to verify?
- Would we want opt-out rights for some content if similar controls arrived in Australia?
- Do we have a way to document abrupt visibility drops after major search changes?
Those are not only visibility questions. They are governance and service-access questions too.
Space for better design
Importantly, this is not a story about shutting AI down. It is a story about shaping it.
The CMA is not saying Google must stop using AI in search. It is saying that if AI is reorganising visibility, speaking in summary form, and affecting who gets found, then that power has to come with responsibilities.
For those building in mental health, this is a useful signal. The future is unlikely to belong to systems that simply automate answers. It is more likely to belong to systems that can show their sources, respect boundaries, allow challenge, and operate in ways people can understand.
Preparing for what comes next
If you work in mental health, digital care, or health information, this ruling is worth watching closely.
A few practical reflections stand out:
- Watch how Google implements attribution in AI Overviews. Source visibility matters deeply in health communication.
- Follow whether complaint channels become meaningful or merely procedural.
- Treat publisher opt-outs as more than a publishing issue; they are also about consent, context, and how AI is allowed to reuse sensitive informational labour.
- Expect the Australian conversation to sharpen if UK implementation shows that these rules are both feasible and useful.
Why the UK ruling matters clinically, not just commercially
For clinicians and service designers, the CMA's ruling is a reminder that the first step into care is increasingly mediated by a handful of private systems. Youth and adult studies already show that people often search online for symptoms and self-diagnosis before they speak to a professional, which means search results shape what they disclose, how urgent they believe their situation is, and whether they seek help at all.
By pulling AI Overviews inside formal conduct requirements, the UK has effectively said that this first, AI-shaped layer cannot be treated as an unaccountable experiment. It must be fair, explainable, and open to challenge when it systematically sidelines trusted services or amplifies lower-quality sources. For mental health, that opens the door to more deliberate collaborations between regulators, search providers, and health systems about what high-stakes queries should prioritise: crisis lines, local services, public clinical guidance, and clearly attributed lived-experience resources.
The law itself does not specify clinical standards or safety templates. It does, however, create space for mental-health organisations to argue that when AI sits at the gateway to care, design choices about ranking, attribution, and opt-out are not only competition questions. They are part of how a society governs access to support, dignity, and timely treatment.
Looking ahead
The UK's Google ruling reflects a broader shift in how societies are responding to AI. The question is no longer only whether AI can summarise, recommend, or rank. The question is whether those systems remain open to scrutiny when they shape public understanding and access to care.
For the mental-health field, that is a significant development. It suggests a future in which digital systems are judged not only by speed and convenience, but by whether they preserve context, support trust, and remain accountable to the people whose lives they touch.
Search is still changing. AI Overviews will continue to evolve. But the UK has now made one thing clear: when AI mediates visibility and understanding at scale, it cannot remain entirely beyond explanation.
Sources
ACCC. (2019, July 25). Digital platforms inquiry - final report. Australian Competition and Consumer Commission. https://www.accc.gov.au/about-us/publications/digital-platforms-inquiry-final-report
ACCC. (2025, May 29). ACCC Digital Platforms Inquiry. Australian Competition and Consumer Commission. https://www.accc.gov.au/about-us/news/social-media/transcripts/accc-digital-platforms-inquiry
ABC News. (2025, October 14). Businesses that rely on Google for customers see dramatic changes. https://www.abc.net.au/news/science/2025-10-15/how-business-adapt-google-pivot-ai-in-search/105867240
Al Jazeera. (2026, June 17). UK orders Google to improve transparency for search services. https://www.aljazeera.com/news/2026/6/17/uk-orders-google-to-improve-transparency-for-search-services
Competition and Markets Authority. (2026a, June 2). CMA secures fairer deal for publishers and improves Google search services in UK. GOV.UK. https://www.gov.uk/government/news/cma-secures-fairer-deal-for-publishers-and-improves-google-search-services-in-uk
Competition and Markets Authority. (2026b, June 16). Google search fair ranking conduct requirement. GOV.UK. https://www.gov.uk/find-digital-markets-measures/google-search-fair-ranking-conduct-requirement
Competition and Markets Authority. (2026c, January 27). CMA proposes package of measures to improve Google search services in UK. GOV.UK. https://www.gov.uk/government/news/cma-proposes-package-of-measures-to-improve-google-search-services-in-uk
GOV.UK. (2026, June 16). Further CMA action to secure a fairer deal for businesses and improve Google search services. https://www.gov.uk/government/news/further-cma-action-to-secure-a-fairer-deal-for-businesses-and-improve-google-search-services-in-uk
Reuters. (2026, June 17). UK orders Google to improve search transparency to boost competition. https://www.reuters.com/world/uk-regulator-sets-out-conduct-requirements-googles-search-services-2026-06-17/
Sky News. (2026, June 3). Google search given nine months to implement UK conduct requirement. https://news.sky.com/story/google-search-given-nine-months-to-implement-uk-conduct-requirement-13550313
The Guardian. (2026, June 3). What does the UK watchdog's new Google AI results rule mean for publishers? https://www.theguardian.com/business/2026/jun/03/what-does-uk-watchdog-new-google-ai-results-rule-means-publishers
Related reading on Mindful Machines Journal: At The Threshold, Algorithmikē Psychē, Can AI Care? The Future of Mental Health and Machine Empathy.
Comments
Post a Comment