AI search is becoming an operating surface for brand teams. The practical question is not whether a single model response changed—it is whether the change reveals something your team can verify and improve.
What the CMA proposed
On September 23, the UK Competition and Markets Authority published updated proposals for Google’s general-search user-choice requirement. The proposal would require Google to show Android and Chrome users a choice of search services at first use, then prompt them annually to choose a default. Its new element is that qualifying AI assistants could be included on those screens, subject to technical and security criteria. The CMA explicitly names services such as ChatGPT and Perplexity as examples of the changing search market. This is a proposal open for consultation, not a decision that puts any assistant on a choice screen today.
Why the default setting belongs in the discovery discussion
A default service can influence which product a person opens when they begin researching a category, comparison, or purchase. That does not determine the answer they receive, the sources shown, or the brands recommended. It does shape an early access point. The CMA’s own strategic-market-status decision noted that default positions are important routes to general search and that ChatGPT and Perplexity could be set as Chrome defaults only through extensions rather than appearing on the current Android choice screen. For brand teams, that makes this a distribution and user-acquisition signal—not evidence of a new AI-ranking factor.
Attribution is part of the proposal, too
The updated CMA proposal would also require search providers featured on choice screens to attribute publisher content fairly, so users can reach source material and understand where results came from. That language matters in an answer-led interface, where an answer, a link, and an underlying source are related but not interchangeable. The proposal does not prescribe a universal citation design, publish a complete technical standard for every assistant, or prove that a particular publisher will receive more visits. It sets a direction for the providers that may qualify for the UK choice mechanism.
This is not a mandate for Google to list every AI assistant
The CMA says an assistant would need to meet relevant technical and security requirements. It is consulting on the updated proposal until October 9 and expects a final decision by the end of 2026. Neither the press release nor the consultation establishes a final eligibility list, a launch date, the ordering of any choice screen, or the share of users who would switch services. Reuters similarly reported the plans as proposals. Any statement that ChatGPT, Perplexity, or another assistant will definitely become a Google default in the UK would get ahead of the evidence.
What it does and does not change for AI-search visibility
The consultation does not announce a change to Google Search ranking, AI Overviews, AI Mode, Gemini, ChatGPT retrieval, Perplexity citations, or how a named brand is described. It does not mean a site will be cited more frequently or that referrals will rise. If implemented, broader access to an assistant could eventually change the mix of discovery surfaces a UK audience uses. That is a market-structure possibility, not a measurable outcome yet. Treat a platform’s default position, an answer mention, a visible citation, a referral, and a conversion as separate records.
The publisher context is more concrete than a generic “AI regulation” story
The choice-screen proposal follows the CMA’s June publisher conduct requirement for Google Search. That existing requirement includes controls over use of search content in generative-AI features, clearer information about that use, engagement metrics, and steps toward clear attribution and access to content. The September consultation does not replace those measures; it extends the user-choice discussion as AI assistants enter the search-provider frame. The practical takeaway is narrow: publisher access and attribution are now being addressed both inside Google’s AI-search obligations and in proposals about which services users can select.
A restrained operating response for brands and publishers
Do not change content or media plans because of a consultation. Instead, keep a clean baseline for the UK: the buyer prompts that matter, the answer systems currently used, answer and source captures, referral records, and a clear account of market and device conditions. Publishers should continue to make claims, authorship, product facts, and source pages easy to inspect regardless of the entry point. Brand teams should retain coverage across more than one answer surface, because a default-setting proposal neither makes one assistant universal nor removes the need to understand how each system frames a category.
What remains unknown
The CMA has not made its final choice requirement, and Google has not announced the implementation details that would follow a final decision. Public evidence does not yet show which assistants will meet the criteria, how the screen will be presented, whether users will switch, or whether a different default changes publisher traffic or commercial discovery. The consultation also cannot resolve the separate questions of source selection, answer quality, recommendation behavior, and downstream value. Those will still require dated, user-facing observation and first-party data.
Bottom line
The CMA has proposed treating qualifying AI assistants as possible search choices on Google Android and Chrome in the UK, while tying the idea to clearer publisher attribution. That is a meaningful acknowledgement that AI assistants can be routes into search and discovery. For now, it remains a proposed access mechanism rather than a platform rollout, ranking change, or visibility guarantee. Keep monitoring the answers and sources people actually see; revisit the distribution question if and when the CMA reaches a final decision.