Bot Protection: Using CAPTCHA to Prevent AI Scraping

Debevoise has a very good blog on AI and other emerging technologies, and the firm posted an announcement about an action that it has recently taken to protect its content from AI bots looking to scrap data from its blog. Here’s an excerpt:

If you’ve visited the Debevoise Data Blog recently, you will have noticed that before you can read a post, you now have to click through a CAPTCHA. We debated whether to add this small bit of friction, but ultimately decided that it was necessary. Over the past few months, we noticed a sharp rise in traffic to our blog from what looks like training or retrieval by large language models. We are strong believers in the value of AI and proud of the content that we publish here. So, while we are happy to make it freely available to clients, students, academics, and policymakers, we are also running a business, and we need to take reasonable steps to limit the ability of machines to harvest and repurpose our work, for easy use without attribution, by others who are offering competing services.

The firm says it is monitoring traffic to its blog to see if this step achieves its objective, and will update readers on what it finds.