Publishers Are Hijacking Your Back Button Because AI Search Destroyed Their Traffic — Now Google Is Cracking Down

You have experienced this. You click a link from Google, read the page, and hit the back button to return to search results. Except you do not go back. The page reloads. Or redirects to another article. Or shoves a full-screen ad in your face. The back button, one of the most fundamental navigation tools in web browsing, simply stops working.

This is back-button hijacking. And Google has had enough.

On April 13, 2026, Google published a spam policy update that named back-button hijacking explicitly as a "malicious practice" and gave publishers 60 days to stop. The deadline is June 15. After that, sites caught manipulating browser history will face manual actions and ranking penalties. Google AdSense is also eliminating the back-button trigger for vignette ads by the same date.

But here is the real story: publishers are not doing this because they are evil. They are doing it because AI search has systematically destroyed their referral traffic, and they are running out of options.

What Is Back-Button Hijacking?

Back-button hijacking is a technique where websites inject entries into your browser history to prevent you from leaving. When you press the back button, instead of returning to the previous page (typically Google search results), the site redirects you to another page on their domain, shows an interstitial ad, or reloads the current page.

The mechanics vary. Some sites use JavaScript to push multiple history entries. Others use redirect chains that create a loop. The result is always the same: the user is trapped.

Taboola and Outbrain, the two largest content recommendation platforms, offer back-button hijacking as a toggle option for publishers. It is not hidden. It is a feature. Publishers enable it because it generates additional ad impressions on the way out, monetizing users who were already leaving.

Barry Adams of Polemic Digital called it what it is: "a terrible, dark UX pattern designed to deceive or manipulate users." He is right. But the question nobody is asking is: why have so many publishers turned to this pattern in the last 18 months?

The Answer: AI Search Killed Their Traffic

Google's AI Overviews now appear on roughly 40% of searches and reach 2.5 billion monthly users. When Google answers a question directly in the search results, there is no click. The user gets what they need from the AI-generated summary and moves on. The website that would have received that visit gets nothing.

This is zero-click search, and it has been growing for years. But AI Overviews have accelerated the trend dramatically. Publishers who relied on informational query traffic, the kind where someone asks a question and clicks the first result, have been hit hardest. That traffic is now being answered by AI before a click ever happens.

The numbers tell the story. According to SparkToro's annual zero-click search analysis, the percentage of Google searches that end without any click to an external website has been climbing steadily. AI Overviews have pushed this trend into overdrive.

Publishers feel this in their revenue. Less traffic means fewer ad impressions. Fewer ad impressions means less revenue. Less revenue means layoffs, cutbacks, and desperation.

The Desperation Cycle

What we are seeing is a cycle that goes like this:

Step 1: Google changes its search product. AI Overviews, featured snippets, knowledge panels, and now AI Mode all reduce the need for users to click through to websites.

Step 2: Publisher traffic drops. Not gradually, but in sharp declines that correlate directly with AI search feature expansions.

Step 3: Publishers get desperate. They look for any way to squeeze more revenue from the visitors they still get.

Step 4: Publishers turn to dark patterns. Back-button hijacking, aggressive interstitials, autoplay videos, fake chat widgets, forced app installs. Anything to capture one more ad impression.

Step 5: Google cracks down. Because these patterns degrade the user experience, Google penalizes the sites using them.

Step 6: Publishers lose even more traffic from the penalties. Return to Step 3.

This is a doom loop. And it is being driven by the structural shift from click-based search to answer-based search.

Google's Enforcement Playbook

Google's April 13 spam policy update was unusually specific. The company does not usually name individual tactics in policy documents. But back-button hijacking got the explicit callout, along with a 60-day compliance window ending June 15.

The policy states that manipulating browser history to prevent users from navigating away from a page is a violation of Google's spam policies. Sites found in violation will receive manual actions, which can result in demotion or removal from search results entirely.

Separately, Google AdSense announced on May 8 that it will stop serving vignette ads triggered by the back button after June 15. This removes the financial incentive for one of the most common forms of hijacking.

These are good moves. Back-button hijacking is a genuinely hostile user experience. But enforcement treats the symptom, not the disease.

Why Enforcement Is Not Enough

Google cracking down on dark patterns is like a city banning tents while ignoring homelessness. The underlying problem remains.

Publishers are not hijacking back buttons for fun. They are doing it because the economics of web publishing have been upended by AI search. When Google answers questions directly, websites that spent years building content to answer those same questions lose their primary distribution channel.

Consider the economics. A publisher that previously received 100,000 visits per month from Google might now receive 60,000. That 40% drop in traffic translates directly to a 40% drop in ad revenue. Fixed costs like editorial staff, hosting, and content production have not dropped 40%. Something has to give.

For many publishers, that something is user experience. They sacrifice UX to monetize the remaining traffic more aggressively. It is a rational response to an irrational situation. It is also self-defeating, because Google penalizes the behavior, which reduces traffic further.

The Real Solution: AI Visibility

The publishers who will survive the transition to AI search are not the ones who find better ways to trap users. They are the ones who build AI visibility.

AI visibility means getting your content cited by AI search engines. When ChatGPT, Google AI Overviews, Perplexity, or any AI-powered search product answers a user's question and cites your brand as a source, you get attribution even without a click.

This is the new currency. Not clicks, but citations. Not traffic, but visibility in AI-generated answers.

The publishers investing in generative engine optimization (GEO) are the ones positioning themselves for the post-click economy. They are building content architectures that AI engines can parse and cite. They are implementing structured data that makes their content legible to AI models. They are creating original research and data that AI engines prefer to cite because it adds credibility to their answers.

Meanwhile, publishers still fighting the click economy war are doubling down on dark patterns and watching their traffic and rankings decline simultaneously.

What This Means for Your Brand

If you are a brand that depends on search traffic, whether you think of yourself as a publisher or not, the back-button hijacking crackdown is a warning sign. The old model of optimizing for clicks and monetizing pageviews is eroding.

The brands that adapt will focus on three things:

Build citation authority. Create content that AI engines want to cite. Original data, expert analysis, unique perspectives. Content that makes AI answers better.

Implement structured data. Make your content machine-readable. Schema markup, llms.txt files, clear content hierarchies. Give AI engines every reason to cite you.

Measure AI visibility. Stop measuring success only in clicks. Track whether your brand appears in AI-generated answers. Track which engines cite you and which do not. Track how your citation presence changes over time.

The Bigger Picture

The back-button hijacking story is really a story about what happens when a market structure changes faster than the participants can adapt. Google moved from directing users to websites to answering questions directly. Publishers built their entire business model on the old structure. The gap between the old model and the new reality is where dark patterns emerge.

Google is right to enforce against hijacking. Users should not be trapped on websites. But until the underlying economic pressure is addressed, publishers will keep finding new ways to extract value from declining traffic. The next dark pattern is already being developed somewhere.

The only sustainable path forward is building visibility in the new system. AI search is not going back to the click model. The question for every brand and publisher is whether you will be cited by the AI engines answering your customers' questions, or invisible to them.

The crackdown deadline is June 15. But the real deadline was months ago. The click economy has been dying since AI Overviews launched. The brands that recognized this early and invested in AI visibility are already ahead. The ones still optimizing for clicks are running out of time, and running out of back buttons to hijack.