Sanjay B Consulting

Google Search Can Now Draw Its Own Pictures, and Finally Put a Number on Canonical Delays

Featured

I go through Google's own announcements and documentation every week, because this is where the real signal lives before anyone writes a hot take about it. This week's biggest story is AI Overviews image generation, Google's new ability to draw a picture instead of linking to yours, and it changes what I'm telling clients to do about their visual content right now. A second update, on canonical fix timelines, changes what I tell them about technical SEO.

Here's my breakdown, and what I'm doing differently because of it.

AI Overviews Image Generation Is Here

Google Images turned 25 this week, and Google marked the anniversary with two changes I'd call structural.

AI Overviews can now generate custom images from a text prompt, built on Google's Nano Banana model, the same image system Google has been pushing across Search and Chrome all year. Google is also rebuilding the Images homepage into a personalized, real-time gallery that updates based on your interests, with saved collections showing up as tabs above the feed. Both features roll out over the coming weeks.

Here's why I'm flagging this for brands. AI Overviews already answer a huge share of queries without sending a single click to a website. Now that same surface can satisfy an image search entirely on its own, generating a picture instead of linking to anyone's photography, product shots, or visual content.

If your brand relies on image search for discovery, this is the moment to invest in visual assets Google genuinely can't replicate with a prompt, real product photography, original diagrams, branded visuals with context a generated image won't have. It's also worth tightening the structured data and surrounding context on your existing images, so there's a clear reason for Google to send someone to your page instead of just generating a substitute.

Google Puts a Number on Canonical Fix Delays

Google added new language to its canonicalization troubleshooting guide this month, and it finally gives me something concrete to tell a client.

The guide now states that pages can stay in a duplicate cluster for up to two weeks after you fix the underlying content issue, and that pages split out faster when the new content is clearly distinct from the rest of the cluster. That two-week window covers content fixes specifically. Redirect issues, canonical tag corrections, and server misconfigurations are treated as separate problems with their own resolution paths. I've had this exact conversation with clients more times than I can count. "I fixed the content, why hasn't Google noticed?" Now I have Google's own number to point to.

Two weeks is the outside edge, not a guarantee, and it doesn't cover every canonical scenario. But it turns an open-ended wait into a real, documented expectation, and that's worth a lot when you're the one explaining the timeline to a client or a boss.

Now, these two updates sit at different layers of how Search actually works, but they point at the same direction.

These two updates sit at different layers of how Search actually works, but they point at the same shift. With AI Overviews image generation pulling more weight as a content generator now, not just a summarizer, real pressure lands on any brand whose visibility depends on being found rather than replicated. At the same time, Google is getting more transparent about the mechanics underneath the surface, which makes it easier to set honest expectations with clients instead of guessing.

If I had to boil it down to one line for my clients:

Your visual strategy needs to earn its place now, and your technical fixes finally come with a receipt.

Let's Talk About Where Your Content Stands

I am tracking this for my existing clients navigating AI Overviews, GEO, and AI search visibility. If you're also trying to figure out whether your brand is positioned to survive an AI Overviews image generation substitute , or you just want a straight answer on how long a canonical fix should realistically take, let's talk.

If you want the deeper dive on how AI search visibility works beyond this one update, I broke it all down in my guide to AI search visibility and GEO.

Reach out through contact@sanjayb.com, or drop an email to me at sanjay.bhattacharya@primotech.com 

Sanjay B.
Sanjay B.

With over 16 years at the forefront of strategic business growth, Sanjay Bhattacharya collaborates with CEOs and founders to reshape market positioning and drive sustainable success. Throughout his journey, he has worn many hats—from Fractional CMO for fast-growing startups to serving as Head of Marketing & Business Strategy at PRIMOTECH. He has been Featured in Under30CEO, American Marketing Association, CMO Times, CTOsync, DesignRush, Earned, HubSpot, MarketerInterview, and more.

Posted on July 20, 2026.
AI OverviewsAI Overviews image generationAI search visibilitycanonicalizationGeoGoogle Images updatenanoNano Bananatechnical SEO