Skip to main content

WikiHow Sues OpenAI: Did AI Steal Our How-To Guides?

WikiHow, the internet's beloved repository of step-by-step guides—from fixing a leaky faucet to acing a job interview—has filed a lawsuit against OpenAI, the company behind ChatGPT. The suit, lodged in the U.S. District Court for the Southern District of New York, alleges that OpenAI scraped more than 11,000 WikiHow articles without permission, using them to train its GPT models. This, WikiHow argues, infringes on at least 1,200 of its registered copyrights.

The core of the dispute? OpenAI's models, after being fed WikiHow's vast trove of practical wisdom, can now generate similar guides on demand. Instead of clicking through WikiHow's pages, users can simply ask ChatGPT for instructions, and it will produce a tutorial that closely mirrors the original content. For WikiHow, this isn't just a nuisance—it's a direct hit to their traffic and revenue.

In the complaint, WikiHow paints a stark picture: AI-generated content is replacing the work of human contributors, slashing the time and effort needed to produce similar guides. This, they argue, erodes the incentive for creators to keep producing high-quality articles. The lawsuit seeks not only financial compensation but also an injunction to stop OpenAI from using WikiHow's content in the future.

OpenAI, for its part, has responded with a familiar defense: their models are trained on publicly available data, and they believe this falls under the "fair use" doctrine. They argue that using web content to train AI is akin to a human reading and learning from publicly accessible information. It's a stance that has been tested in other lawsuits, but the outcome here could set a significant precedent.

This case is more than a legal spat between two tech entities. It's a flashpoint in the ongoing debate about the boundaries of AI training and copyright law. As AI becomes more integrated into our daily lives, questions about who owns the data that powers it—and who gets paid when that data is used—are becoming impossible to ignore.

For content creators, this lawsuit is a bellwether. If WikiHow wins, it could embolden others to seek compensation for their work being used in AI training. If OpenAI prevails, it might greenlight more aggressive scraping practices, potentially reshaping the economics of online content.

Either way, the outcome will likely influence how AI companies approach data sourcing and how platforms like WikiHow protect their intellectual property. It's a legal battle that could define the future of content creation in the age of artificial intelligence.

Key Points

  • WikiHow's Allegations: OpenAI scraped over 11,000 articles, infringing on at least 1,200 copyrights.
  • Impact on WikiHow: AI-generated guides divert traffic and revenue, undermining creators' incentives.
  • OpenAI's Defense: Claims fair use, arguing training on public data is legal.
  • Broader Implications: The case could set a precedent for AI training and copyright law, affecting creators and tech companies alike.