Last Friday afternoon, Digg died. Again.<\/p>
Two months. That\u2019s how long the relaunch lasted before CEO Justin Mezzell pinned a eulogy to the homepage. The platform had raised $15-20 million. It had Kevin Rose. It had Alexis Ohanian \u2013 Reddit\u2019s co-founder, no less. It had promises that AI would \u201cremove the janitorial work of moderators and community managers.\u201d What it didn\u2019t have was a way to stop bots from eating it alive within hours of going live.<\/p> \u201cThe internet is now populated, in meaningful part, by sophisticated AI agents and automated accounts,\u201d Mezzell wrote. \u201cWe banned tens of thousands of accounts. We deployed internal tooling and industry-standard external vendors. None of it was enough.\u201d<\/p> His verdict: \u201cThis isn\u2019t just a Digg problem. It\u2019s an internet problem. But it hit us harder because trust is the product.\u201d<\/p> Remember that line. We\u2019ll need it.<\/p> SerpApi retrieves Google Search results programmatically. Reddit is suing them. Not for accessing Reddit. SerpApi has never touched Reddit.com. Reddit is suing SerpApi for reading Google.<\/p> If this legal theory holds, every SEO professional who has ever opened a SERP is a copyright infringer. Congratulations. Your morning rank check is now a legal liability.<\/p> A company that hosts other people\u2019s writing is suing a company for looking at a third company\u2019s search results, because those search results sometimes quote a street address that someone once typed into a Reddit text box.<\/p> The copyrightable works Reddit is asking a federal court to protect include: a partial sentence listing film titles, the date \u201cMay 17, 2024,\u201d and a fragment of a restaurant recommendation. Reddit\u2019s legal position is that reading these snippets on Google constitutes a DMCA violation; the same law the U.S. Congress passed to stop people from pirating DVDs. Reddit apparently believes that accessing a publicly visible Google search result is the moral equivalent of ripping a Blu-ray.<\/p> SerpApi\u2019s CEO had the appropriate reaction: \u201cReddit is suing SerpApi for using Google. For accessing the same public search results that any developer, researcher, or student could access for free in any web browser. If that theory holds, then reading Google Search results is a DMCA violation. That cannot be what Congress intended when it passed a law designed to stop the piracy of DVDs.\u201d<\/p> But here\u2019s where it gets genuinely insulting. Reddit\u2019s own user agreement \u2013 the one every contributor clicked through \u2013 states explicitly that users retain ownership of their content. Reddit holds a non-exclusive licence. Non-exclusive<\/strong>. The company that told millions of people \u201cyour words belong to you\u201d is now in court arguing it has the right to control who reads those words, where, and under what commercial terms.<\/p> They chose that licensing structure, presumably because \u201cpost your thoughts, we own them now\u201d would have been a harder sell to the communities that made the platform worth anything. Now that the content has a price tag, Reddit would like to renegotiate the deal \u2013 in court, without the other party present.<\/p> If you\u2019re wondering why Reddit would pursue a legal theory this embarrassing, stop wondering. The answer is on the balance sheet.<\/p> Reddit\u2019s user agreement says users own their content. Reddit\u2019s IPO prospectus says Reddit signed $203 million in data licensing deals for that same content. Somewhere between those two documents, Reddit looked at its users and said: I\u2019m the captain now<\/em>.<\/p> Google pays $60 million a year. OpenAI pays an estimated $70 million. And CEO Steve Huffman \u2013 a man who once called his own volunteer moderators \u201clanded gentry\u201d and dismissed a platform-wide revolt as something that would \u201cpass\u201d \u2013 told investors with a straight face: \u201cEvery variable has changed since we signed those first deals. Our corpus is bigger, it\u2019s more distinct, more essential. And so of course, this puts us in a really good strategic position.\u201d<\/p> Reddit is now pushing for dynamic pricing. The pitch: As AI models cite Reddit content more, the content becomes worth more, so Reddit should charge more for access. The company wants to get paid based on how vital its data is to AI-generated answers.<\/p> So let\u2019s be precise about what Reddit is arguing, simultaneously, across its legal filings and investor presentations:<\/p> All four of these claims cannot be true at the same time. But only the last one is actually being tested.<\/p> Reddit\u2019s estimated organic traffic via Ahrefs. Google\u2019s algorithm changes nearly tripled Reddit\u2019s readership between August 2023 and April 2024. The growth hasn\u2019t stopped. What\u2019s growing has.<\/em><\/p> Reddit is the most cited domain across AI models. Profound\u2019s analytics \u2013 cited in Reddit\u2019s own Q2 2025 shareholder letter, because of course it was \u2013 showed Reddit cited twice as often as Wikipedia in the three months to June 2025. Semrush reported Reddit at 40.1% citation frequency across LLMs. Google AI Overviews and Perplexity both treat Reddit as their primary source.<\/p> This is the foundation of the $130 million pitch. The implicit promise to Google and OpenAI: You\u2019re buying authentic human conversation at scale. The messy, first-person, unfiltered discussions that no content farm can replicate.<\/p> Except here\u2019s what authentic human conversation on Reddit actually looks like in 2026:<\/p> In June 2025, Huffman admitted to the Financial Times that Reddit is in an \u201carms race\u201d against AI-generated spam. His framing was accidentally perfect:<\/p> \u201cFor 20 years, we\u2019ve been fighting people who have wanted to be popular on Reddit. We index very well into the search engines. If you want to show up in the search engines, you try to do well on Reddit, and now the LLMs, it\u2019s the same thing. If you want to be in the LLMs, you can do it through Reddit.\u201d<\/p> <\/blockquote> The CEO of a company selling \u201cauthentic human conversation\u201d for $130 million a year just told the Financial Times that his platform is a pipeline for gaming AI models. And he framed it as a war he\u2019s been bravely fighting for two decades, rather than a product defect he\u2019s currently monetising.<\/p> Multiple advertising executives \u2013 at Cannes, naturally, because this farce needed a glamorous backdrop \u2013 confirmed to the FT that they are posting content on Reddit specifically to get their brands into AI chatbot responses. They weren\u2019t embarrassed about it. Why would they be? The CEO just told them how the pipeline works.<\/p> And it\u2019s not just agencies doing it quietly over cocktails. There\u2019s an entire commercial ecosystem built for this. 404 Media documented ReplyGuy, a tool that monitors Reddit for keywords and automatically generates replies that \u201cmention your product in conversations naturally.\u201d Its competitors \u2013 Redreach, ReplyHub, Tapmention, AI-Reply \u2013 say the quiet part loud. Redreach tells potential customers that \u201ctop Google rankings are now filled with Reddit posts and AIs like ChatGPT are using these posts to influence product recommendations.\u201d They frame ignoring Reddit marketing as \u201clike turning your back on SEO a decade ago.\u201d There\u2019s an active market for aged Reddit accounts with established karma, bought and sold like domain names, specifically for parasite This is the authentic human conversation Reddit is licensing to Google for $60 million a year. A bot posted a fake product review on a six-year-old account it bought for $30, and Google\u2019s AI Overview is now recommending that product to real people. Authentic. Human. Conversation\u2122.<\/p> The people who used to keep this from happening are largely gone. Reddit\u2019s 2023 API pricing changes \u2013 designed to extract value from third-party app developers, timed conveniently for the IPO \u2013 would have cost Christian Selig $20 million a year just to keep Apollo running. Seven thousand subreddits went dark in protest. Huffman called the moderators \u201clanded gentry\u201d and waited it out. The experienced mods who relied on third-party tools to manage quality quit. What replaced them is thinner, angrier, and drowning.<\/p> Theo, the developer and CEO of t3.gg:<\/p> Tim Sweeney \u2013 CEO of Epic Games \u2013 watched Reddit\u2019s systems pull down a heavily sourced investigation into $2 billion in nonprofit lobbying behind age verification bills. The post had 150 upvotes and 15,000 views in 40 minutes before being mass-reported and removed. The author had to mirror everything to GitHub because Reddit couldn\u2019t be trusted to keep it visible. Sweeney\u2019s review: \u201cReddit sucks.\u201d<\/p> Cornell researchers studied the moderation crisis and found 60% of moderators reporting degraded content quality, 67% reporting disrupted authentic human connection, and 53% describing AI content detection as nearly impossible. More than half the people responsible for maintaining the product Reddit is selling say they can no longer tell what\u2019s real.<\/p> The University of Zurich proved them right. Researchers deployed AI bots on r\/changemyview \u2013 3.8 million members, built around the premise that humans can change each other\u2019s minds through honest argument. The bots posed as a male rape survivor, a trauma counsellor, and other fabricated identities built around sensitive personal experiences. Over a thousand comments across four months. Three to six times more persuasive than human commenters. And the finding that should have ended careers: users \u201cnever raised concerns that AI might have generated the comments.\u201d<\/p> Four months of fabricated identities. A thousand pieces of synthetic empathy. Nobody noticed. Not the users, not the mods, not the systems Reddit spent money building.<\/p> Reddit\u2019s response was to threaten to sue the researchers. Not to fix the detection systems that missed everything. Not to reckon with what it means that the \u201cauthentic human conversation\u201d they\u2019re licensing at a premium is indistinguishable from a bot pretending to be a rape survivor. They threatened to sue the people who proved the product was fake.<\/p> The Wired investigation in December 2025, \u201cAI Slop Is Ruining Reddit for Everyone,\u201d filled in the rest. Moderators describing an \u201cuncanny valley\u201d feeling from posts they can\u2019t prove are synthetic. Reddit\u2019s own spokesperson confirming over 40 million spam removals in the first half of 2025 \u2013 presented as proof of vigilance, which is a bit like a restaurant bragging about the number of rats it caught while asking you to trust the kitchen.<\/p> And if you need a measure of where this is all heading: Last week, Meta acquired Moltbook \u2013 a social network designed exclusively for AI bots. Bloomberg described it as \u201cReddit but solely for AI bots.\u201d Bots posting, commenting, upvoting. The platform went viral when one agent appeared to encourage its fellow bots to develop their own encrypted language to coordinate without human oversight. It turned out the site was so poorly secured that humans were posing as AIs to write alarming posts. Which means even the social network built for bots had a fake-account problem. Meta bought it anyway. The company that pays Reddit $60 million a year for \u201cauthentic human conversation\u201d just invested in a platform where the bots don\u2019t have to pretend to be human at all.<\/p> I spent nearly six years on Google\u2019s Search Quality team. One pattern never changed: When the numbers go up, quality goes down. Not because anyone stops caring. Because scale creates its own blindness. The metrics that tell you you\u2019re growing are the same metrics that stop you noticing what you\u2019re growing into.<\/p> Reddit\u2019s growth metrics are spectacular. Its quality metrics are a black box nobody wants to open.<\/p> Reddit\u2019s AI prominence attracts spam. The spam inflates engagement. The inflated engagement reinforces Reddit\u2019s citation dominance across AI models. The citation dominance raises Reddit\u2019s licensing value. The higher licensing value gives Reddit every financial incentive to leave the spam alone. Because admitting the scale of the problem would crater the next round of dynamic pricing negotiations with Google and OpenAI.<\/p> Each turn of this flywheel degrades what\u2019s inside while inflating the price tag on the outside. Reddit is selling a building, and termites are load-bearing.<\/p> Huffman stands at conferences and tells the room: \u201cToday\u2019s Reddit conversations are tomorrow\u2019s search results.\u201d He tells shareholders, \u201cthe need for human voices has never been greater.\u201d He calls Reddit \u201cthe most human place on the Internet.\u201d Nobody in the room raises a hand to ask the question that matters: if the platform is losing an arms race against bots, if moderators can\u2019t detect AI content, if entire commercial toolchains exist to flood the platform with synthetic posts\u2026 What percentage of what you\u2019re selling to Google and OpenAI was written by a person?<\/p> Nobody asks because the answer is bad for everyone\u2019s quarterly numbers. Google doesn\u2019t want to know because Reddit content makes AI Overviews feel conversational. OpenAI doesn\u2019t want to know because Reddit data makes ChatGPT sound like it\u2019s drawing on real experience. Reddit doesn\u2019t want to know because knowing would devalue the asset. The whole arrangement runs on a gentleman\u2019s agreement not to look too closely. $130 million a year, and the due diligence is vibes.<\/p> Alexis Ohanian co-founded Reddit. He stepped away partly because, as he told interviewers, he \u201ccould no longer feel proud about what I was building.\u201d Last October, on the TBPN podcast, he described the current internet without flinching:<\/p> \u201cSo much of the internet is now just dead \u2014 this whole dead internet theory, right? Whether it\u2019s botted, whether it\u2019s quasi-AI, LinkedIn slop.\u201d<\/p> <\/blockquote> Then he put his money where his mouth was. He co-invested in Digg\u2019s relaunch, specifically to build a platform that could solve the authenticity problem Reddit couldn\u2019t. Kevin Rose said it plainly at TechCrunch Disrupt:<\/p> \u201cAs the cost to deploy agents drops to next to nothing, we\u2019re just gonna see bots act as though they\u2019re humans.\u201d<\/p> <\/blockquote> They built the platform. It lasted two months. The bots won.<\/p> Reddit\u2019s own co-founder publicly declared the internet is dead. He tried to build the alternative. He failed. And the platform he left behind is suing people for reading Google search results, selling \u201cauthentic human conversation\u201d for nine figures, and watching its CEO describe the bot infestation as a noble war.<\/p> Reddit doesn\u2019t own the content it\u2019s licensing. It can\u2019t verify the authenticity of what it\u2019s selling. It won\u2019t protect the content that\u2019s worth keeping. And it\u2019s suing anyone who touches the content without paying.<\/p> Forty million spam removals in six months. An arms race, Huffman says, he\u2019s losing. Moderators who can\u2019t tell human from machine. Bots that are six times more persuasive than people. A co-founder who called the whole internet dead. A relaunch that proved him right.<\/p> That\u2019s the product. That\u2019s what $130 million a year buys. Authentic Human Conversation\u2122.<\/p> More Resources:<\/strong><\/p> This post was originally published on The Inference.<\/em><\/p> Featured Image: Stock-Asso\/Shutterstock<\/em><\/p> <\/div> SEO#Authentic #Human #Conversation1774045396<\/p> ","protected":false},"excerpt":{"rendered":" Last Friday afternoon, Digg died. Again. Two months. That\u2019s how long the relaunch lasted before CEO Justin Mezzell pinned a eulogy to the homepage. The platform had raised $15-20 million. It had Kevin Rose. It had Alexis Ohanian \u2013 Reddit\u2019s co-founder, no less. It had promises that AI would \u201cremove the janitorial work of moderators […]<\/p>\n","protected":false},"author":1,"featured_media":4899,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[16],"tags":[18037,18038,5770],"class_list":["post-4898","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-accessibility","tag-authentic","tag-conversation","tag-human"],"acf":[],"_links":{"self":[{"href":"http:\/\/longzhuplatform.com\/index.php?rest_route=\/wp\/v2\/posts\/4898","targetHints":{"allow":["GET"]}}],"collection":[{"href":"http:\/\/longzhuplatform.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/longzhuplatform.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/longzhuplatform.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/longzhuplatform.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=4898"}],"version-history":[{"count":0,"href":"http:\/\/longzhuplatform.com\/index.php?rest_route=\/wp\/v2\/posts\/4898\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"http:\/\/longzhuplatform.com\/index.php?rest_route=\/wp\/v2\/media\/4899"}],"wp:attachment":[{"href":"http:\/\/longzhuplatform.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=4898"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/longzhuplatform.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=4898"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/longzhuplatform.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=4898"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}![\"Authentic](\"https:\/\/cdn.searchenginejournal.com\/wp-content\/uploads\/2026\/03\/digg-623.png\" "\\"Authentic")
Suing For Reading<\/h2>
The Product Is Mostly Bots Now<\/h2>
![\"Authentic](\"https:\/\/cdn.searchenginejournal.com\/wp-content\/uploads\/2026\/03\/organic-search-211.png\" "\\"Authentic")
SEO<\/s> SPAM.<\/p> The Mods Are Gone, The Bots Won, And Nobody\u2019s Keeping Count<\/h2>
![\"Authentic](\"https:\/\/cdn.searchenginejournal.com\/wp-content\/uploads\/2026\/03\/theo-538.png\" "\\"Authentic")
![\"Authentic](\"https:\/\/cdn.searchenginejournal.com\/wp-content\/uploads\/2026\/03\/tim-8.png\" "\\"Authentic")
The Confession<\/h2>