Google now tells sites to manually fact-check and review all AI-generated content for accuracy and trustworthiness before publishing — metadata included.

On Oct. 1, Google updated its guidance on using generative AI content by adding three sentences to the "Focus on accuracy, quality, and relevance" section. The additions clarify that generative models predict likely word sequences rather than retrieve facts, and call it "critical to manually fact-check and review all AI-generated content for accuracy and trustworthiness before publishing." Google also revised the metadata line to show the review process applies to titles, meta descriptions, structured data, and image alt text. According to Google's documentation changelog, the changes bring the page in line with presentations at developer events.

The update puts AI-generated metadata under the same manual checks as article drafts. A review process focused only on main content might overlook other fields that Google says can appear in Search results.

Gemini Links Carry UTM Tags, And Google's Mueller Sees Them

A Reddit user found UTM parameters on links from Google's Gemini to websites, and Google's John Mueller replied that he sees them too. Mueller asked whether Gemini passes a referrer that sites can already track, then offered to forward details about the referrer's handling to the team, preferably with examples he can reproduce.

Roger Montti reports that Google has not documented the tagging process, so it is unclear when Gemini adds the parameters. Tagged links give analytics a way to identify Gemini visits without depending on the referrer. Comparing sessions with UTM tags against referrer-based Gemini data shows how many visits each method tracks.

Desktop CTR Fell In Q2 While Mobile Rose, AWR Data Shows

A Q2 2026 clickthrough report from Advanced Web Ranking shows desktop CTR down at the top organic positions and mobile rates up — the reverse of its Q1 findings. Across all searches, desktop CTR at positions one and two decreased by a combined 5.27 percentage points from Q1 to Q2, while mobile CTR at these positions rose by a combined 6.59 points. The figures combine changes at two positions and are not tied to any specific site.

AWR says its tables, based on its CTR tool, source data from Search Console. The comparison period spans the end of a Search Console logging error that misreported impressions from May 13, 2025, to April 27, 2026. The report does not mention this error.

That omission matters for analysts. Any Q1-versus-Q2 comparison in Search Console data includes April 27, the date Google reports the impressions error was fixed. Since CTR divides clicks by impressions, that date could be a point where measurement shifted — before anyone reads a CTR change as a shift in user behavior.

AI Mode Information Monitoring Rolls Out Globally

Robby Stein from Google announced on X on Sept. 28 that AI Mode's information monitoring is now rolling out globally to all users. Google AI Ultra and Pro subscribers had already been using the feature. All of Stein's examples focused on local topics and shopping.

Users tell AI Mode what to monitor, and Search continuously scans for updates. The feature also suggests monitoring tasks during searches. The post describes "info monitoring capabilities" rather than information agents, a feature introduced for Ultra subscribers in June.

Stein's June post said updates would include links to the web, but neither it nor the Sept. 28 post explains how Search picks the sources behind an update. Once the rollout reaches an account, setting up a task for a given type of business shows what an update looks like and whether it carries links.

Judge Mehta Dismisses Penske And Chegg AI Overviews Suits

U.S. District Judge Amit Mehta granted Google's motions to dismiss the antitrust lawsuits filed by Penske Media and Chegg over AI Overviews. In his Sept. 30 ruling covering both cases, Mehta determined the plaintiffs failed to plausibly allege an agreement in which Google exchanged search traffic for their content. He stated they lacked legal standing to challenge Google's alleged maintenance of its search monopoly, and he found that Penske did not establish that Search and AI Overviews are separate products.

The judge chose not to consider the state unjust-enrichment claims and said the court does not treat alleged harms lightly. Mehta wrote that antitrust law isn't a substitute for lawmakers dealing with the economic upheaval new technology causes. The ruling does not order Google to change how it uses publisher content in Search or AI Overviews.

What's Written Down

The week's thread is documentation. Google logged its fact-checking guidance change in a changelog. Gemini's UTM tags surfaced in a Reddit thread; the monitoring rollout arrived via an executive post on X. The AWR report omits the Search Console error linked to its comparison window, and Mehta's ruling spelled out what antitrust law requires a complaint to allege. Documented changes can be integrated into workflows directly. For undocumented ones, evidence — a Gemini-tagged session, a monitoring task you set up — has to be built by hand.