Contents

Key Takeaways 1

1. Introduction and Summary 2

2. Data Portability Conduct Requirement 3

3. Publisher Conduct Requirement 4

4. Fair Ranking Conduct Requirement 4

5. User Choice Conduct Requirement 5

Recommendations 5

Conclusion. 6

1. Introduction and Summary

On October 10, 2025, the UK’s Competition and Markets Authority (CMA) designated Google as having strategic market status (SMS) in the market for general search services using its authority under the Digital Markets, Competition and Consumers Act 2024 (Act).[1] On January 26, 2026, the CMA issued proposed conduct requirements (PCRs) to be imposed upon Google’s general search services, which specify how Google “must conduct itself in relation to the digital activity in respect of which it has SMS.”[2]

The PCRs encompass four broad conduct requirements. First, a data portability requirement that “give[s] consumers who use Google’s general search services the right to port their data to other businesses.”[3] Next, a publisher conduct requirement which “provide[s] publishers with informed and meaningful choice over how Google uses their content, crawled for general search, for the purpose of generative AI.”[4] Third, a fair ranking conduct requirement with the aim to ensure that “publishers are treated fairly where their content is used in general search.”[5] Finally, a user choice conduct requirement “to enable consumers to make informed and effective choices and promote greater competition among search providers.”[6]

The Information Technology and Innovation Foundation (ITIF) appreciates the opportunity to comment on the PCRs. ITIF commends the CMA for issuing detailed guidance on the PCRs but is concerned that they do not satisfy the proportionality criteria set forth in the CMA’s Digital Markets Competition Regime Guidance by virtue of imposing unnecessarily burdensome requirements that will harm UK consumers and innovation.

The structure of this comment mirrors the structure of the PCRs: first, a discussion of why the data portability conduct requirement is disproportionately harmful; second, an analysis of how the publisher conduct requirement is not the least restrictive alternative by virtue of extending into AI markets; third, an assessment of how the fair ranking conduct requirement is far more onerous than what is necessary to address anticompetitive behavior; and fourth, an examination of why the user choice requirement is unlikely to have any significant effect toward helping rival search engines. Recommendations and a brief conclusion follow.

2. Data Portability Conduct Requirement

The CMA explains that the “aim for our intervention is to ensure that UK consumers who use Google’s general search services can effectively port their data to other businesses to develop new services or otherwise share the value of that data” pursuant to the “fair dealing” objective in section 19(6) of the Act.[7] While, as the CMA describes, Google offers a voluntary application programming interface (API) for data portability, its “provisional assessment is that this uncertainty risks negatively impacting investment decisions, and therefore growth and innovation”; indeed, the CMA notes that “Google submitted that the incremental costs of formalizing the API are likely to be negligible, given that Google already voluntarily provides the API.”[8]

However, the assessment of whether this PCR does not produce disadvantages which are disproportionate to the aim does not adequately consider the innovation effects of mandating Google’s data portability API in the manner contemplated. While Google already provides data portability tools, a mandatory regime would constrain its flexibility in how such tools are designed, scoped, and evolved over time. Unlike a voluntary framework, a mandatory obligation may reduce incentives to invest in platform-specific features whose value can be immediately and frictionlessly replicated by rivals without equivalent investment. Mandated portability may also incentivize firms to compete by optimizing around the extraction and reuse of incumbent-developed user data, rather than investing in independent data generation or improvements to core search technology.

3. Publisher Conduct Requirement

With its publisher PCR, the CMA attempts to satisfy not just the fair dealing objective in section 19(6) of the Act but also the trust and transparency objective in section 19(8) of the Act.[9] The CMA envisions Google “implementing a new control allowing publishers to opt out of their Search Content being used in the grounding of content in its search generative AI features”; indeed, as the CMA notes, “Google-Extended already enables publishers to control whether their Search Content is used in training Google’s foundation models and grounding Google’s broader generative AI services,” such that it should be “technically feasible” for Google to develop similar controls for AI search features.[10]

It is far from clear that the publisher PCR is the least onerous of equally effective options to achieve the Act’s aims. Specifically, rather than addressing behavior by Google that protects its general search services position for which it was designated as having SMS, the CMA is seeking to prevent Google from competing in next-generation AI markets—well outside the scope of what the CMA should be focusing on. In fact, there is no competitive problem in the nascent and dynamic AI search space that the Act needs to solve. For example, OpenAI’s chief financial officer recently stated that AI is “blowing open the search markets,” noting that the company doubled its search share from roughly 6 percent to 12 percent in just 6 months.[11]

4. Fair Ranking Conduct Requirement

Like the publisher PCR, the fair ranking PCR is similarly justified by the CMA in terms of fair dealing as well as trust and transparency.[12] This PCR requires both that “Google’s relevant ranking decisions should be based on objective and non-discriminatory factors that provide users with the most relevant and high-quality results to their specific query” and that Google provide information about its search rankings that is “sufficiently detailed to allow publishers to understand how ranking decisions are made and to identify possible instances of discrimination or distortion.”[13]

Despite the CMA’s arguments to the contrary, it is far from clear that the fair ranking PCR is no more onerous than necessary to achieve the Act’s aims. Indeed, rather than focus on instances of where Google may have engaged in anticompetitive self-preferencing, the fair ranking PCR’s non-discrimination provision applies much more broadly to any search ranking—even if Google’s own products and services are not implicated and it thus lacks any prima facie incentive to leverage its purported market power in general search services. What’s more, a blanket ban on discrimination is not only bound to have negative welfare effects for consumers by limiting the ability for content that may be more desirable for a user to compete for better placement, but may be undermined by the transparency requirements that open the door to publishers “cheating” such that their content features prominently instead of that which consumers actually want to see.

5. User Choice Conduct Requirement

Finally, the CMA’s user choice PCR, which is designed to further the open choices objective in section 19(7) of the Act, requires that “Google shall provide Users, or ensure that Users are provided, with a Search Choice Screen that cannot be skipped on all Relevant Browsers and all Relevant Devices that enables Users to select a Default Search Provider.”[14] For the CMA, appropriately designed choice screens are “key to ensuring that users can make active and informed choices” about their use of a search engine.[15] This design is detailed at length by the CMA to satisfy criteria associated with greater coverage, balanced prompts, and more.[16]

The user choice PCR is highly unlikely to be effective in achieving its aim. In connection with antitrust relief imposed in Europe, Google was forced to implement similar search choice screens that allowed Android users to choose between Google and other search engines. Rather than create opportunities for competitors, not only did consumers not appreciate the worse user experience, but the choice screens had virtually no effect on Google’s search market share—simply put, users chose Google because it has the best search engine.[17]

Recommendations

For these reasons, ITIF has concerns with the PCRs and respectfully offers the following recommendations for the CMA to consider:

▪ Data Portability Conduct Requirement: Although the CMA identifies the benefits that could flow from its data portability PCR in terms of increased investment, relatively little attention is paid to the investment disadvantages that will result from forced sharing both with respect to Google’s incentives to innovate, as well as those of third parties. The CMA should thus reassess whether speculative concerns about decreased investment from Google withdrawing its API outweigh the real disincentives to invest that would be created by mandating Google’s API. 

▪ Publisher Conduct Requirement: While the PCRs are ostensibly designed to address Google’s SMS status in general search services, the publisher PCR attempts to limit Google’s ability to compete in dynamic, next-generation search markets that require no regulatory intervention by the CMA. Rather than overly burden Google with unnecessary measures, the CMA should focus solely on PCRs that target the general search services market where it was deemed to have SMS.

▪ Fair Ranking Conduct Requirement: The CMA’s decision to impose a broad fair ranking PCR opts for heavy-handed regulation of the general search market rather than addressing the self-preferencing behavior that would in theory involve Google leveraging market power in general search to gain an advantage in a second market. To minimize false positives, the CMA should tailor its PCRs to address conduct that can constitute anticompetitive behavior.

▪ User Choice Conduct Requirement: Rather than put forward a user choice PCR that might meaningfully address Google’s purported dominance in general search services, the CMA doubles down on a well-worn remedy that has demonstrably not achieved significant procompetitive benefits while degrading user experience. As such, the CMA should drop this PCR and go back to the drawing board to determine how the Act can actually be used to improve user choice.

Conclusion

Having implemented the Act, the CMA faces a fundamental choice as to how it will approach the regulation of digital platforms: faithfully apply its proportionality analysis in a way that focuses on effective, necessary, unburdensome, and on balance pro-innovation conduct requirements, or simply roll out the standard fare of remedial measures that have long been thrown at firms like Google but which have dubious merits at best. Amidst this time of increasing technological dynamism and global tensions, and given the special relationship that exists between the United States and the UK, the CMA should reassess how it can implement the Act in a more light-touch way that ensures not just that British innovation and consumers will be made better off, but that Britain will avoid being subject to potential retaliatory actions leveled at jurisdictions that implement digital antitrust regulations that target America’s leading technology companies.

Endnotes