Google AI Mode and ChatGPT Are Both Testing Bottom Citation Cards

Both Google and OpenAI moved quietly on the same front this week. According to Google AI Mode & ChatGPT Tests Bottom Card Citations, reported by Barry Schwartz on July 22, 2026, Google AI Mode and ChatGPT are each running tests that relocate citation cards away from inline positions and push them toward the bottom of the answer. Two platforms, two slightly different approaches, one shared direction — and a big open question about what it means for the publishers and brands whose content is being cited.

The Google AI Mode variant was spotted by Sachin Patel, who shared a video on X. According to Schwartz's report, Patel wrote: "Google appears to be testing a change to citation cards in AI Mode. When opening a free listing product, the citation cards move to the bottom of the page." The mechanism anchors the user directly to the citation section at the bottom rather than keeping sources visible inline alongside the answer.

The ChatGPT variation was caught by Brodie Clark via his SERP Alert account. Clark noted that ChatGPT is experimenting with how it links out to sources — loading a citation box as an overlay when a user clicks, rather than presenting it in the flow of the response. Schwartz also reported that the organic shopping experience within ChatGPT now features a new display and product panel that appears on click.

What's Actually Changing

Right now, most users of Google AI Mode see citation cards surfaced inline or near the top of a response — close to the claim being supported. Moving them to the bottom is a significant layout shift. It changes the visual hierarchy of the page and, critically, changes when a user sees a source.

In Google's test, the anchor behavior is the key mechanic. Rather than scrolling manually, users get jumped directly to the citation card. That sounds helpful, but it also means the user is finishing the AI answer first and only then encountering your brand as a source. The path to your site just got a little longer.

ChatGPT's overlay approach is different. It keeps the citation off the main reading path entirely until someone deliberately clicks to see it. That's a more intentional interaction — which could mean fewer but more motivated clicks. Or it could mean most users never bother. We don't have click data yet, and the source doesn't provide any. But the directional signal is clear: both platforms are reducing the visual weight of citations.

Schwartz's closing question in the piece cuts right to it: "Will these interfaces lead to more or less clicks?" That's not rhetorical. It's genuinely unanswered — and it's the only metric that actually matters for your site.

What This Means for Your SEO and AI Visibility

Picture this: you've worked hard to get your content cited in AI Mode. Your brand shows up. Your URL appears. But now that citation is sitting at the bottom of a long AI-generated answer that fully resolves the user's question before they ever scroll. That's the scenario these tests are pointing toward.

Getting cited by an AI has always been a partial win. You get brand impressions, but the traffic conversion from citation to click has been inconsistent at best. This UI direction suggests both Google and OpenAI may be deprioritizing the click-out experience in favor of keeping users engaged within the AI interface. That's a problem if your business model depends on referral traffic from search.

At the same time, there's a counterargument worth sitting with. An anchor jump to a citation card at the bottom of an AI answer might actually surface your brand more prominently than a small inline footnote that most users skip. If the card design is rich enough — your brand name, a summary, a clear link — it could convert better. We just don't know yet.

What I've seen consistently over the past couple of years is that AI citation visibility and traditional organic ranking are becoming two separate games. You can rank on page one and get zero AI mentions. You can get heavily cited in AI Mode and see organic impressions drop. Tracking both independently is no longer optional — it's how you understand where your actual search presence lives.

What to Do Now

This is still a test. Neither change has been confirmed as a permanent rollout. But the direction is consistent with where both platforms have been heading, so treating this as a signal rather than waiting for a full launch is the smarter move.

1. Start tracking AI citations separately from organic rank. If you're not already monitoring whether your content appears in AI Mode or ChatGPT responses, you're flying blind. AI visibility tracking gives you a baseline to measure against when interfaces like this shift.
2. Audit your citation-worthy content. Both platforms cite content that is structured, authoritative, and specific. If your pages are vague or buried in fluff, they won't get cited — and even if they do, a bottom-card UI makes weak brand presentation a bigger liability. Tighten your headings, lead with your clearest answer, and make sure your brand name appears early in the content.
3. Strengthen your brand signal in citations. If your domain or brand name isn't immediately recognizable, a bottom-of-page card is easy to skip. Build brand search volume through other channels so users recognize your name when they see it in a citation card.
4. Watch your referral traffic from AI sources. Pull Google Analytics or Search Console data and segment for AI-driven referrals where possible. Establish a benchmark now so you can detect changes when these tests roll out more broadly.
5. Don't panic-optimize. These are UI tests. The underlying citation logic — what content gets cited and why — has not been reported as changing. Keep producing structured, expert-level content. That's still the entry ticket.

Background and Context

This isn't the first time Google has experimented with how citations are surfaced in AI responses. Since AI Overviews launched, there has been ongoing testing around source card size, placement, and visibility. The general trend has been toward cleaner AI answer surfaces with citations as secondary elements rather than primary navigation.

ChatGPT's evolution on citations has followed a similar arc. Early versions of the product showed sources prominently at the end of every response. Over time, the interface has become more selective about when and how sources are displayed — and this overlay test is another step in that direction.

The broader pattern here is one I keep coming back to at conferences: AI platforms are optimizing for user experience within the interface, not for publisher referral traffic. That's not a conspiracy — it's just product design. But it means SEOs and content marketers need to think about AI citation as a brand channel, not just a traffic channel. The click may become the exception rather than the rule.

If you want a simple way to start keeping tabs on where your brand is being cited across AI platforms, Aergos has an AI visibility checker that can help you see where you stand. It's a useful starting point while these interfaces continue to evolve.