The videos are all over social media, making an irresistible offer to students: go ahead and let artificial intelligence do your homework — with the latest technology, you won’t get caught. If you hate writing, you can avoid it. Even established ed-tech companies are marketing with a wink and a nod.
These kinds of tutorials are now pervasive on TikTok and YouTube. They show students how to use tools known as humanizers and autotypers, which make it easier than ever to cheat. The videos — sometimes labelled ads, sometimes not — target college and high school students.
Humanizers rewrite AI-produced text to make it sound less robotic, formulaic and trite. Autotypers slowly drip words and sentences into documents, making it appear as if papers were typed at a human pace when in fact, they were produced by AI. They even fabricate typos, deletions and revisions. Both tools can help students evade software designed to detect AI.
Colleges and K-12 schools are trying to keep up, with AI detection becoming a significant expense. But educators attempting to restrict the technology, worried about students failing to develop basic skills, are often lagging in what tech-industry leaders are calling a detection arms race.
In some cases, the same companies selling detection tools are also making apps that allow students to cheat, including by writing papers for them or rephrasing text. The apps promise to help them avoid accusations of misconduct by scanning their work before they submit it, allowing them to rewrite passages identified as AI. Jenny Maxwell, head of education at Superhuman, the AI company that makes Grammarly, called the race between detection and evasion “ultimately, a dead end”.
Even before AI chatbots, the Internet had made cheating easier, in part through the simple mechanism of copy-and-paste plagiarism. Now, the landscape is more complex.
A recent College Board survey of professors in the US found three-quarters reported their students were using AI to write, and more than 90 per cent of respondents were concerned about plagiarism and dishonesty. Many institutions have seen a sharp increase in student disciplinary cases for academic misconduct, much of it related to the use of AI.
OpenAI’s ChatGPT and Google’s Gemini are the most popular AI tools among students. But just beneath these behemoths is a roiling, fiercely competitive market of legacy ed-tech purveyors and tiny startups, all using social media to tell young people that their academic lives could be easier — much easier — if they embrace AI. Some startups explicitly teach students how to cheat.
Meanwhile, established companies often urge students to use their tools responsibly as aids for studying, research, brainstorming, outlining and revision. But many of them are simultaneously producing technology that can easily be used to plagiarise and cheat.
Smaller companies are sometimes more direct. In one TikTok video, Carter Smith, a young tech influencer known as CarterPCs, gleefully shows viewers how an autotyping and humanizing app called Grubby AI can make it seem like a person naturally wrote an essay that was, in fact, produced by ChatGPT.
Autotypers are a response to the fact that many teachers and professors now check a document’s version history for signs of AI use. If 1,000 words suddenly appeared in a Word or Google document at 11.59pm, it could mean the student pasted in text produced by a chatbot. GrubbyAI and its many competitors are finding ways around those systems.
Dripwriter’s website says the app provides “believable typos and fixes” along with “background autotyping so your essay keeps working when you step away”. Duey.ai describes itself as the “#1 autotyper for Google Docs”, tells customers that when they’re too tired or busy to focus, or out with friends, “The document looks like you wrote it.”
It is a crowded market, in which upstarts constantly pop up. A TikTok video about another app, Typeflo, told students that they could relax, watch YouTube and eat a sandwich while their essays were produced for them. Typeflo was registered to Daniel Huddleston, a professor at Emory University’s medical school. His response: the app was developed and marketed by his teenage son, and he had not been fully aware of its social media presence.
Some professors are increasingly concerned about Grammarly, an app that has existed for 17 years as a sort of muscular spell-check. It now offers an “authorship” tool that helps professors screen for AI misconduct, by analysing a document’s version history. At the same time, the app allows students to generate writing from scratch, humanise text, and scan and replace phrases that could set off AI detectors.
Like other AI executives, Maxwell of Superhuman said cheating has always existed but represents only a small segment of student AI use. “I can’t solve the human behaviour issue that is cheating or pushing the easy button,” she said. “It is out of our realm.”
NYTNS