Here is a scary thought for anyone trying to find the truth on the internet. Google’s AI just lied to a law professor about a paper he never wrote. It didn’t just get the details wrong; it made up a whole fake article with a fake title to sound smart. This isn’t a glitch; it’s a feature of how these machines work, and it puts us all at risk.
Professor Albert Alschuler from the University of Chicago found this out the hard way on September 19, 2026. He wanted to find a real response from Professor Bradley Smith to an old debate they had. Instead of a link to a real document, Google’s AI Mode gave him a summary of a ghost story. The machine claimed Alschuler wrote a piece called “The Ancient and Honorable Art of Truculent Rhetoric.” He never wrote it. The AI just cooked it up out of thin air.
The Machine Makes Stuff Up
Alschuler asked the AI if he had replied to Smith, and the bot said yes with absolute confidence. It even gave him a fake title and a summary of points Alschuler supposedly made about campaign finance. The AI claimed Alschuler argued that the Supreme Court’s definition of corruption was “dangerously insufficient.” While that might be what Alschuler thinks, the bot invented a specific article to prove it. The real article exists, but it has a different name.
This happens because the AI is designed to be helpful, not truthful. It sees a pattern and fills in the blanks with whatever sounds right, even if it’s a complete lie. It told Alschuler he addressed Smith’s critiques point-by-point in a piece that doesn’t exist. The bot was so convincing it made the user believe he had published something he didn’t. This is the “Ancient and Honorable Art” of AI hallucination, and it is spreading fast.
If a law professor can’t trust his search results, what happens in a New York City classroom? The Heritage Foundation warns that families and policymakers need to get a handle on this mess fast. They point out that technology can be a “valuable instructional tool when it supports strong pedagogy and effective teaching practices,” but only if we know how to use it. Right now, we are flying blind. Delaware lawmakers are studying the issue, but their report won’t come out until 2029.
That is too slow for New York. Kids are using these tools right now to write essays and do homework. If the AI lies about facts, students learn lies. The Heritage Foundation argues that local school boards need to step up and make rules. They say students should only use AI for schoolwork when a teacher specifically allows it. We can’t wait for federal bureaucrats to figure this out while the next generation learns to trust a machine that makes things up.
As Kirsten Baesler, assistant secretary for elementary and secondary education, put it, “The U.S. Department of Education believes these decisions [about education technology] are best made by states and local communities.” New York City needs to take that advice and act now. We need training for teachers to spot these fake citations before they become part of a student’s grade.
The Bottom Line
We love this city, and we want our schools to be the best in the world. But we can’t let a computer program write our history or our law. When the AI tells you it found a source, check the source. If you can’t find it, the machine is lying. This isn’t about hating technology; it’s about demanding honesty from it. If we don’t draw a line, we are just handing our education over to a glitchy robot that thinks it knows everything.
This is opinion and commentary from the RapForMayor team.
