Google just handed eligible publishers a new control that’s going to trigger way too much dramatic Slack chatter.

The Search generative AI control lets eligible site owners decide whether their content can appear within Google’s generative AI search features. Those surfaces include AI Overviews, AI Mode, and generative AI features in Discover.

The important part is what the control doesn’t do. It doesn’t remove the website from Google’s traditional search results.

If somebody treats this like a full Google opt-out switch, they’re already solving the wrong problem.

The New Reports Measure Visibility, Not Traffic

The smarter move is to stop guessing and start measuring. Google introduced dedicated Search Console generative AI performance reports for Search and Discover.

As of August 31, 2026, Google says these reports are available to websites worldwide.

The reports show:

• Impressions from generative AI features
• Pages appearing within those features
• Visibility by country
• Device data for Search
• Trends by hour, day, week, or month

That’s useful, but it isn’t a complete traffic report. The dedicated reports don’t currently show clicks, CTR, queries, average position, or conversions.

They tell you whether and where your pages appeared. They don’t prove that AI visibility produced a visit, lead, or sale.

So yes, the old “AI is stealing all our traffic” argument is getting harder to hide behind. However, “AI visibility is driving valuable customers” can’t be proven from this report alone either.

The data is a starting point, not a verdict.

A Blanket Opt-Out Isn’t Automatically the Smartest Move

For many websites, immediately opting out could mean abandoning another source of brand visibility before understanding its value.

That doesn’t make every opt-out decision lazy or irrational. Publishers may have legitimate concerns involving licensing, content control, business models, or how their work is summarized. Those decisions can extend beyond SEO performance.

The mistake is making a major change without documenting the reason or establishing a baseline.

Before changing the setting, review:

1. Which pages receive generative AI impressions
2. Which countries and devices produce that visibility
3. Whether exposure is rising or falling
4. Which page types appear most often
5. How overall organic traffic and conversions changed during the same period
6. Whether the visibility supports the brand’s publishing and revenue goals

Search Console’s AI reports can’t directly connect impressions to conversions. Teams will need to compare those trends with broader Search Console data, analytics, leads, sales, and branded-search behavior.

That comparison still won’t prove perfect causation. It will produce a better decision than guessing.

Test Page Types Instead of Treating the Site as One Blob

A sitewide control can create different consequences across different sections.

Informational articles may receive substantial AI visibility without producing many direct visits. Product pages, service pages, original research, and branded resources may behave differently.

Review those groups separately. Don’t assume a policy article, buying guide, local service page, and original study contribute the same kind of value.

Track the pages that appear most often, then compare their broader performance over time. Look for changes in organic clicks, branded searches, assisted conversions, leads, and revenue.

If the analysis becomes a reporting mess, an experienced search visibility partner can help connect AI impressions with the business metrics that actually matter.

Use the Control and the Data Together

This isn’t the moment for chest-thumping declarations about opting in or out. It’s a moment for cleaner measurement, sharper page-level analysis, and less theater.

Establish the baseline. Review the affected pages. Define what success means. Document the business risks. Then decide whether participation supports the site’s goals.

Google has given eligible publishers more control and everyone more visibility data. The strongest decisions will come from teams that understand the limits of both before clicking anything dramatic.