New UK CMA Publisher Conduct Requirement for Google Search
An ISOC LIVE Summary
Sources: Competition and Markets Authority - UK competition regulator; Alissa Cooper - Executive Director, Knight-Georgetown Institute; Theo Bamber - Chief Executive, News Media Association; Sajeeda Merali - Chief Executive, PPA (Professional Publishers Association)
The UK Competition and Markets Authority (CMA) issued its first formal conduct requirement under the Digital Markets, Competition and Consumers Act, targeting Google Search and the use of publisher content in generative AI systems and AI-enhanced search features.
The requirement follows the CMA’s designation of Google as holding Strategic Market Status in general search services. The regulator argued that publishers increasingly depend on Google for traffic, discovery, and advertising visibility, while Google’s rapid integration of generative AI into search products risks weakening publisher bargaining power and reducing referral traffic.
The CMA identified three core concerns during its consultation process:
Publishers lacked meaningful choice over how their content was used in AI systems.
Publishers lacked transparency regarding when and how their content was used.
AI-generated search experiences risked weakening attribution and reducing traffic back to original sources.
The CMA framed the conduct requirement as an attempt to establish baseline fairness and transparency obligations while preserving user access to search services and supporting long-term sustainability of high-quality online content.
Publisher Controls Over AI Usage
The finalized conduct requirement obliges Google to provide publishers with controls allowing them to prevent their search content from being used for specified generative AI purposes.
The final version expanded the draft proposal by explicitly covering both AI training and fine-tuning activities. Publishers must also be able to restrict the use of their content for “grounding,” meaning the use of publisher content to generate or support live AI responses.
For grounding within general search, publishers must be able to apply controls at domain, directory, and page level. The CMA emphasized that these controls should be genuinely functional and should not require publishers to sacrifice participation in ordinary Google Search indexing.
This represents a major shift from the earlier “all-or-nothing” dynamic in which publishers seeking to avoid AI usage risked losing visibility in conventional search results.
The CMA also said publishers should be able to exercise these controls without suffering discriminatory treatment in search ranking or indexing.
Transparency and Reporting Requirements
Google must provide publishers with clear and accessible explanations describing how publisher content may be used in generative AI systems and AI-enhanced search experiences.
The conduct requirement additionally obliges Google to provide reporting and metrics related to user engagement involving publisher content within AI-generated search interfaces.
The CMA stated that publishers require sufficient information to assess the commercial impact of AI search experiences on audience reach, referral traffic, and monetization.
The regulator also indicated that transparency obligations will need to evolve as Google introduces new AI search products and interfaces.
Attribution and Access to Original Sources
A major focus of the conduct requirement concerns attribution.
Google must take reasonable steps to ensure that publisher material appearing within AI-generated search experiences is clearly attributed and linked back to original sources.
The CMA acknowledged concerns that AI summaries and synthesized answers could weaken incentives for journalism and professional publishing if users consume summaries without visiting source websites.
The requirement therefore aims to preserve visibility for original reporting, analysis, academic publishing, and other forms of professionally produced content.
Industry and Policy Reactions
The News Media Association welcomed the conduct requirement as a significant step toward “levelling the playing field” between publishers and dominant digital platforms. Theo Bamber said the effectiveness of the regime would ultimately depend on implementation and enforcement as Google’s AI search products evolve.
Earlier in the consultation process, the PPA (Professional Publishers Association) also supported stronger publisher protections. Sajeeda Merali argued that publishers should have meaningful control and transparency regarding how their content is used in generative AI systems, while preserving their ability to participate in ordinary search indexing.
Both organizations emphasized that the relationship between publishers and AI-driven search systems is becoming increasingly important to the long-term sustainability of journalism and professional publishing.
CMA Framing and Broader Regulatory Context
The CMA presented the measure as an early intervention rather than a complete solution to search market concentration.
The regulator acknowledged that Google’s integration of AI into search is evolving rapidly and that publisher dependency on Google creates substantial asymmetries in negotiating leverage.
The conduct requirement is intended to establish minimum protections while broader investigations into search competition, ranking fairness, user choice, and data portability continue.
The measure also reflects wider international pressure on dominant technology platforms around AI training, publisher compensation, attribution, and transparency.
Alissa Cooper Commentary
Alissa Cooper framed the finalized conduct requirement as a potentially important step for publishers facing declining click-through rates and AI-generated content that may misrepresent original reporting.
Cooper noted that the CMA improved the final requirement after consultation by extending publisher controls over AI training to also cover fine-tuning activities. She connected both that improvement and the unresolved issue of granular feature-level controls directly to recommendations made in the Knight-Georgetown Institute consultation response.
However, Cooper argued that important ambiguities remain unresolved. In particular, she criticized the absence of clear definitions for “training” and “grounding,” warning that the lack of precision could leave room for strategic interpretation or compliance gamesmanship.
She also argued that publishers still lack sufficiently granular controls over individual AI search features and experiences.
Concerns About AI Search Evolution
Cooper connected the timing of the CMA decision to Google’s recent announcements at Google I/O, where the company previewed increasingly AI-centric search experiences blending traditional search, conversational AI, and agent-style functionality.
She questioned whether publishers are effectively receiving opt-out rights at the same moment that AI-generated interfaces may fundamentally reshape or replace the traditional search results page.
In her analysis, the practical value of publisher controls may ultimately depend on whether publishers possess enough bargaining power to negotiate meaningfully with Google.
She warned that smaller publishers without significant commercial leverage may gain little practical benefit from formal opt-out mechanisms alone.
YouTube and Creator Concerns
Cooper also criticized the conduct requirement for failing to address YouTube creators.
She argued that YouTube represents one of the most valuable datasets available for AI model development, yet creators currently lack meaningful control over whether their videos, transcripts, and associated metadata are used in AI systems.
According to Cooper, creators themselves — rather than solely the platform — should have authority over how their work participates in AI development and AI search ecosystems.
Structural Competition Issues
Cooper situated the UK measure within a fragmented international regulatory landscape involving search, AI, data access, licensing, and publisher bargaining.
She referenced parallel activity in the EU around data-sharing and interoperability and emerging US discussions around AI licensing and syndication frameworks.
However, she argued that regulators globally continue to avoid confronting what she considers the core structural source of Google’s market dominance: default search distribution agreements.
According to Cooper, Google’s payments securing default search placement across browsers, mobile devices, and operating systems remain central to its entrenched power in search markets.
Until regulators directly address those distribution arrangements, she suggested, measures focused on AI usage controls and publisher transparency may provide only partial remedies for a broader structural competition problem.
RESOURCES
CMA secures fairer deal for publishers and improves Google search services in UK - official CMA press release announcing the conduct requirement
Google search publisher conduct requirement - the measure page with all official documents
Publisher CR final decision - full CMA decision document (PDF)
Google’s general search and search advertising services - CMA case timeline from SMS designation through consultation
Alissa Cooper commentary - KGI Executive Director’s analysis of the conduct requirement on LinkedIn
A Missed Opportunity to Address Google’s Market Power in Search in the UK - KGI’s critique of the proposed remedies (Mar 2026)
KGI Public Comment on the Proposed Conduct Requirements - the consultation response Cooper references, PDF (Mar 2026)
NMA Responds To CMA Conduct Requirement For Google Search - Theo Bamber’s reaction for the News Media Association
CMA publishes Conduct Requirements for Google - PPA’s consultation-stage response from Sajeeda Merali (Jan 2026)