The Great Blockade Transcript and Summary

An AI chatbot falsely accused publisher Steve Grove of murder, exposing the dangerous consequences of the web's new invisible wall.

[Speaker 1]: It started with a shooting in Minneapolis. This was just a few weeks ago, January 8, 2026. Federal agents were involved, it was a chaotic scene, and naturally, people turned to the internet to figure out who the shooter was. [Speaker 2]: But they didn't just Google it. A lot of people asked AI chatbots. And one of the leading bots came back with a very specific name. It identified the shooter as Steve Grove. [Speaker 1]: Which was a massive problem. Because Steve Grove wasn't the shooter. Steve Grove is the CEO and Publisher of the Star Tribune-the biggest newspaper in Minnesota. [Speaker 2]: He wasn't even at the scene. But the AI said he was the gunman, stated it as a fact, and spread that misinformation to thousands of users in real-time. [Speaker 1]: Now, usually when we hear stories like this, we think, "Okay, the AI broke. It hallucinated." But that’s not actually what happened here. The AI didn't fail because it was broken. It failed because it was working exactly the way it’s been forced to work. [Speaker 2]: Right. The Star Tribune had blocked that AI from reading its articles. So when the bot went looking for the truth, the only door that was locked was the one guarding the actual facts. So it went next door-to social media-scraped a bunch of rumors, and presented them as truth. [Speaker 1]: Steve Grove wasn’t the victim of a glitch. He was a casualty of a war. For the last two years, we’ve been watching the open web get replaced by a "Great Firewall." We’re seeing a massive, invisible wall going up between the companies that report the news and the machines that distribute it. [Speaker 2]: And that wall is forcing a really dangerous choice: we can either fund journalism, or we can have accurate AI. But right now, it looks like we can’t have both. [Speaker 1]: It’s Thursday, February 5, 2026, and you’re listening to The Angle. [Speaker 2]: To understand how Steve Grove ended up being accused of murder by a chatbot, we have to look at the scale of this blockade. It is much bigger than most people realize. [Speaker 1]: Yeah, we tend to think of the internet as this open library. You walk in, you read what you want. But that library is shutting down. [Speaker 2]: Completely. As of last month, January 2026, 79% of the top news sites in the US and UK are now blocking AI training bots. If you look at real-time retrieval-that’s the bots that look up breaking news-71% of top publishers block those too. [Speaker 1]: So the vast majority of high-quality journalism is now totally invisible to the AI models that most people use every day. And the reason for this comes down to a broken deal. For twenty years, the web ran on a very specific social contract: "I’ll let you crawl my website, you scrape my content, and in exchange, you send me traffic." [Speaker 2]: Right. Content for clicks. That was the trade. [Speaker 1]: But the trade is over. [Speaker 2]: It is. The numbers are pretty stark. Data from Playwire shows that 69% of news queries are now "zero-click." That means the user asks the question, the AI gives the answer, and the user never visits the publisher’s site. There was a user study recently that showed Google’s AI Overviews reduced click-through rates from 15% down to 8%. [Speaker 1]: So from a publisher’s perspective, like the Telegraph or the Chicago Tribune, they’re looking at these…

Related coverage: Technology · Popular stories