What changed
Google's newly surfaced Ads Help page describes Review summaries on Search: Google AI aggregates online reviews from across the web and writes short, positive-only commentary directly into eligible Search ads. It applies to English-language Search campaigns on Google.com in the United States, Australia, United Kingdom and India, across supported devices and ad positions; sensitive verticals such as pharmaceuticals are excluded. Google says it serves summaries only when it predicts an ad-performance benefit and that advertisers cannot directly edit or customise the text. The help page is undated; specialist outlets documented it October 24–26, 2026, after months of visible experiments. It does not establish a universal rollout date, an advertiser opt-out or a reporting interface.
Why it matters
Advertisers can now have a platform-written description of customer sentiment embedded in paid ad copy without direct text control. This shifts brand and accuracy governance from ordinary creative review to managing third-party reviews and monitoring what the platform generates. Agencies and ad-management SaaS tools cannot assume that their approved headlines and descriptions are the only words consumers see. The change is particularly relevant to small advertisers whose reputation depends on a handful of external review surfaces. Google's claim of improved ad performance is predictive, not published outcome evidence.
Where the summaries can appear
Google's help page limits eligibility to Search campaigns on Google.com, English in the US, UK, Australia and India, on supported devices and at all ad positions. It excludes sensitive advertiser categories, naming pharmaceuticals as an example. Eligibility is not a promise that every impression will carry a summary: Google says it only shows the asset when its systems predict a benefit.
Google controls the generated words
The AI gathers reviews from across the web and produces positive-only commentary. Google's documentation explicitly says advertisers cannot directly edit or customise the generated text because it represents user feedback. This differs from a manually approved testimonial or normal responsive Search ad asset; it creates a review-quality and attribution problem if the source material is stale, misattributed or disputed.
What is not documented
The Google help page does not identify the review platforms it reads, provide a review-correction mechanism, say whether account-level automated-asset opt-outs cover this format, or describe impression/performance reporting. Do not infer that the feature can be disabled merely because other automated assets have switches. No evidence establishes that advertisers can choose which positive reviews are used.
Evidence and rollout limits
Google's own help page establishes the feature and rules but has no publication date. Search Engine Roundtable and Search Engine Land reported the new documentation October 26, while PPC Land traced sightings of related experimental review formats since 2025 and highlighted the missing control details. This is a documented feature, not proof that every eligible account already receives it. Google has not released independent uplift measurements.
Operational response for advertisers and tooling
Audit live Search ad appearance, brand mentions and review-source accuracy in eligible markets; keep an evidence trail of questionable generated text and contact Google Ads support if a summary misrepresents the business. Agencies should tell clients that Google may add AI-written review commentary outside approved ad copy. Track official Help updates for eligibility, correction and opt-out controls before making promises about control.