cd /news/ai-policy/google-enables-publisher-opt-out-fro… · home topics ai-policy article
[ARTICLE · art-23414] src=letsdatascience.com pub= topic=ai-policy verified=true sentiment=· neutral

Google Enables Publisher Opt-Out From AI Search Features

The UK Competition and Markets Authority (CMA) imposed a conduct requirement obligating Google to let publishers opt out of generative AI search features and prevent content from being used for model fine-tuning. Google began testing a new Search Console toggle with a subset of UK websites, allowing site owners to withhold content from AI features, though current reports show impressions but not clicks or click-through rates. The lack of click data limits publishers' ability to assess the traffic tradeoff of opting out, as the CMA's requirement calls for click-level metrics not yet present in Google's reports.

read4 min publishedJun 6, 2026

This week the UK Competition and Markets Authority (CMA) imposed a conduct requirement that, according to reporting in The Verge and SEJ, obliges Google to let publishers withhold content from AI Search features and from use in model fine-tuning. Google has begun testing a new toggle in Search Console with a subset of UK websites that lets site owners opt out of appearing in generative AI features, The Verge and CNET report. CNET quotes Mrinalini Loew, Google general manager of Search Ecosystem: "Sites that opt out will not receive traffic or impressions from our generative AI features." Google's new AI performance reports in Search Console currently surface impressions for AI features but, per Search Engine Journal, do not yet include clicks or click-through rates. Editorial analysis: Observers should note that the lack of click data limits publishers ability to judge the traffic tradeoff implied by opting out.

What happened

The UK Competition and Markets Authority (CMA) this week imposed a conduct requirement that, according to The Verge and Press Gazette, requires Google to give publishers control to withhold their content from generative AI search features and to prevent that content from being used for model fine-tuning. The CMA also requires clearer attribution in AI responses with links back to source domains, the CMA said in its announcement quoted by The Verge.

Product update (reported)

Google is testing a new control in Search Console that lets participating website owners manage whether their content and links appear in Google's generative AI features, reporting by CNET and The Verge states. CNET quotes Mrinalini Loew, Google general manager of Search Ecosystem: "Sites that opt out will not receive traffic or impressions from our generative AI features." Google told CNET it is starting the test with a subset of UK publishers and is engaging with regulators like the CMA while testing, per CNET.

Reporting on measurement gaps

Search Engine Journal reports that Google's new AI performance reports in Search Console currently show impressions for AI features but do not provide clicks, click-through rates, or a clear separation between AI-driven impressions and organic search traffic. The CMA's conduct requirement, as described in The Verge and SEJ coverage of the CMA documents, explicitly calls for publishers to be provided click-throughs, click-through rates, and data separated from organic search; that specific data is not yet present in the published Search Console reports, SEJ notes.

Editorial analysis - technical context

At a technical level, impressions alone do not reveal user engagement or referral value. Industry practitioners commonly differentiate between surface-level visibility metrics such as impressions and behavioural metrics such as clicks and session downstreams when estimating monetizable traffic. Observed patterns in similar measurement rollouts show that publishers need click-level and referral-path data to estimate revenue impact, advertising attribution, and subscription conversion funnels.

Industry context

Reporting by The Verge and Press Gazette frames the CMA action as a first-of-its-kind regulatory requirement for a search platform, aimed at improving publishers negotiating leverage and transparency. CNET and The Verge report that Google is testing controls in the UK first and intends, according to The Verge, to expand after testing. Industry observers will see this as a regulatory touchpoint that other jurisdictions may study when assessing platform-publisher dynamics.

What to watch

  • •Whether Google adds clicks,** click-through rates**, and a clear separation of AI-feature metrics from organic search in Search Console, as the CMA documentation and SEJ coverage indicate publishers want. - •The scope of the rollout beyond the initial subset of UK publishers and any timelines reported by The Verge or CNET.
  • •How attribution links required by the CMA are implemented in AI Overviews and whether those links generate measurable referral traffic compared with current search snippets.
  • •Any formal statements or guidance from Google that expand on the mechanics and guarantees of the opt-out toggle; if none are produced, reporting may continue to rely on the CMA text and publisher experiments.

Bottom line

This is a regulatory-driven product change with immediate operational consequences for publishers, but current reporting highlights a measurement gap: publishers have an opt-out control in testing, yet the click-level data necessary to quantify the traffic and revenue tradeoffs is not publicly available in Search Console reports at the time of these articles. Industry practitioners tracking publisher analytics and referral attribution should treat the presence or absence of click-level AI metrics as the key signal of how usable the opt-out control will be in practice.

Scoring Rationale #

This combines a regulator-imposed requirement from the UK CMA with a product-level control in Google Search Console. It materially affects publisher measurement and content-negotiation dynamics, but the immediate practitioner impact is limited by missing click-level data.

Practice with real Ad Tech data

90 SQL & Python problems · 15 industry datasets

[Active Search Campaigns by BudgetEasy](/problems/sql/active-search-campaigns-by-budget)

[High CPC Clicks & Poor Landing PagesMedium](/problems/sql/high-cpc-clicks-poor-landing-page)

[Campaign ROAS by Attribution ModelHard](/problems/sql/campaign-roas-by-attribution-model)

250 free problems · No credit card

See all Ad Tech problems

── more in #ai-policy 4 stories · sorted by recency
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/google-enables-publi…] indexed:0 read:4min 2026-06-06 ·