Google's dishonest and insulting AI Mode justification - a teardown
Google finally responded to widespread criticism of its AI-based search tools, and that response is beyond disappointing.
Since Google’s AI Mode rolled out in the United States in May, media outlets and content creators have been screaming bloody murder.
The feature pulls in content from across the web and displays it as a summary, with just a few source links shown off to the side.
Major US news groups like the News/Media Alliance, which represents more than 2,000 publishers, call AI Mode “theft.” It cuts both visitor numbers and advertising income, they say, since users now get the information they want straight from Google’s answer box rather than from publisher sites themselves.
Independent traffic analysis by Similarweb, Pew Research Center, and others report drops in organic search traffic for some news publishers ranging from 30% to 55% since the widespread rollout of AI Overviews and AI Mode. CNN reports a 30% drop; Business Insider, and The Washington Post report >40% to over 50% drops over a multi-year period.
Six out of ten Google searches now end in what experts call “zero click”—users never leave the search results page or visit a publisher’s website at all.
Sites like HuffPost and the New York Post have seen their Google referrals drop by over 50% since the rollout began in early 2025. Small independent sites face even steeper drops.
Publishers fear that making AI Mode the default search tool could take what’s left of the digital publishing business and crater it entirely.
Publishers can’t opt out of AI Mode without disappearing from regular Google search altogether, so for now, if they want to be visible at all, they have no real choice but to allow Google to harvest their content.
Google finally responded to the controversy with this insulting blog post.
The post says that AI Overviews and AI Mode are driving more complex and diverse queries while maintaining stable overall click volume to websites. They claim that AI is improving the “quality” of clicks by sending users to more relevant and engaging content.
The post is a textbook example of slippery evasion, the kind of bullshit you’d normally get from a weasel politician.
What’s wrong with Google’s post
The post claims that total organic click volume has remained “relatively stable” and that average click quality has increased. The post hides Google’s criteria for “relatively stable.”
Google shits all over third-party reports, claiming they’re flawed or based on isolated incidents, but provides no detailed rebuttal or specific evidence to undermine these external analyses. This is a classic “strawman argument” where opposing data are caricatured rather than addressed substantively.
To this, we should all apply Hitchens’s Razor: “Any claim asserted without evidence can be dismissed without evidence.” So dismiss the claim that third-party reports criticizing AI Mode are flawed.
The language Google uses to describe metrics is conspicuously vague and imbued with optimistic qualifiers like “higher quality clicks” and “people are happier with the experience.”
The term “quality clicks” is narrowly defined by Google, focusing on non-bounce traffic. There is a wide range of other potential metrics for the quality of clicks, which Google conveniently ignores.
Another bullshit tactic: Google takes credit for alleged “improvements” in the “quality” of search while attributing the loss of traffic to vague “user trends.” In other words, the good things we’re claiming are all about us; the bad things that are happening have nothing to do with us.
Also: The whole claim is dripping with vagueness. Google says “user trends are shifting traffic to different sites (how much traffic?), resulting in decreased traffic (how much of a decrease?) to some sites and increased traffic to others (how much of an increase?).
Google frames criticism as if it were based on a false binary choice between supporting the web or embracing AI, presenting itself as a hero that champions both simultaneously. This framing downplays the inherent tension caused by AI Overviews that mediate and condense information, fundamentally altering the balance between users, content creators, and the search engine.
The magnitude of this potential shift for publishers demands concrete data and real numbers, not platitudes and weasel-worded shiftiness. The total absence of detailed methodologies or granular data to substantiate key claims mocks publishers and insults them.
Effective, transparent and honest communication answers the questions people have without raising more questions. Google’s blog post does the opposite. Here are the many questions they raised:
“Our data show people are happier with the experience and are searching more.”
Based on what? How are you measuring “happiness”? “How much is ‘more’?”
“Total organic click volume from Google Search to websites has been relatively stable year over year.”
What does “relatively stable” mean, exactly? 3% or 50% — what’s the criteria for this conspicuously vague claim?
“Average click quality has increased.”
By how much?
“We’re actually sending slightly more quality clicks to websites than a year ago.”
Presumably, clicks always rise year over year. How does the growth of clicks compare with previous years’ growth?
“This data is in contrast to third-party reports that inaccurately suggest dramatic declines in aggregate traffic — often based on flawed methodologies, isolated examples, or traffic changes that occurred prior to the roll out of AI features in Search.”
Which third-party reports? All of them? Some of them? Which reports, exactly?
“With AI Overviews, people are searching more.”
What does that mean? They’re coming to the site more? They’re doing more searches per session? What does “searching more” mean, exactly?
“With AI Overviews, people are seeing more links on the page than before.”
What does “seeing” mean here and why? Are you saying that, although there are materially fewer links, people are noticing them more? They’re eyes are resting on them more? Their minds are engaging with the links more? They’re clicking more? Again, what does “seeing” mean?
“More queries and more links mean more opportunities for websites to surface and get clicked.”
By “more opportunities,” are you saying that there would be more clicks if websites did something different? Are you specifically saying lower clicks are the fault of websites?
“User trends” are resulting in “decreased traffic to some sites and increased traffic to others.”
What’s driving user trends? Are you saying that Google Search has little or no effect on “user trends” in the driving of traffic on the web?
“People are increasingly seeking out and clicking on sites with forums, videos, podcasts, and posts where they can hear authentic voices and firsthand perspectives.”
Are you implying that news articles, blog posts, and newsletter posts like on Substack or Medium are less authentic than videos on TikTok and YouTube or than on podcasts? What exactly are you saying with this statement?
“People are also more likely to click into web content that helps them learn more — such as an in-depth review, an original post, a unique perspective, or a thoughtful first-person analysis.”
Are you implying that the reason traffic is dropping on the 2,000 publishers that form the News/Media Alliance is because they’re not helping people learn, don’t have in-depth reviews, don’t have original posts, don’t offer a unique perspective or thoughtful first-person analysis and that’s why their traffic dropped when you rolled out AI search? Is that what you’re saying or is that not what you’re saying?
“Sites that meet these evolving user needs are benefiting from this shift and are generally seeing an increase in traffic.”
Are you saying that before AI-based search, Google Search was not good at meeting user needs? What are you saying with this statement exactly?
(Note: I’ve asked Google to provide the data behind the claims in its blog post, and I’ll share that information with you if they provide it.)
Google’s slippery messaging here combines statistical ambiguity, strawman arguments, vague optimism, victim blaming, and reframing tactics that collectively try to minimize concerns about publisher disintermediation and obscure inconvenient facts, all while maintaining an appearance of benevolent stewardship over the future of the web.
Google: If you’re going to crush the media ecosystem that feeds the AI machine you hope to make billions of dollars from, at least be honest and transparent with your claims. This blog post is a shameful insult.
More From Elgan Media, Inc.
Smart people use AI to get smarter
AI slop is eating the world
When everything is vibing
Where’s Mike? Oaxaca, Mexico!
(Why I’m always traveling.)
It's a paradigm shift and, in order to survive, publications will need to hide their content behind a paywall and AI firms will need to pay up to access it. I believe Reddit and New York Times have such agreements. This is the media's Napster moment.
Brilliant tear down