What Is The 30% Rule For AI?

AI seems to be spreading everywhere. Companies like GitHub provide AI coding tools to make programming easier, and it seems impossible to avoid AI-generated images online. As AI has expanded, guidelines and "rules" for using it have started to pop up online. One of the most notable is the 30% rule for AI.

The rule has a couple of meanings, and how you interpret it might depend on your industry. According to some online reports, this idea may have originated in education, where a program called Turnitin is often used to verify the originality of written work and assess the likelihood of AI use. The 30% rule is not defined by Turnitin, but rather a common interpretation. 

But what does it mean exactly? According to online posts, the 30% rule means that anything you plan to turn in to your professor should score below a 30% threshold for AI-generated work. That means at least 70% of the work that you've done on a paper or essay should be easy to prove as human-based using Turnitin's detection system. However, it isn't as simple as checking the box for being under that 30% threshold, since the threshold itself isn't a universal standard that all professors follow.

A second potential interpretation of the 30% rule in AI is that AI should do 70% of the work, with human efforts making up the remaining 30%. Again, it's less a hard, defined rule, and more a guideline to help workers understand how much to lean on AI while also retaining human intervention in areas where it matters most — issues such as quality control, leadership, ethical judgments, and creative or critical-thinking work.

The goal, then, is to have AI handle repetitive, data-intensive tasks, freeing human workers to focus on what AI can't do reliably. So, instead of just worrying about that AI will start taking jobs, think of AI as a tool for those jobs while still leaving humans to make final decisions.

While these are the primary two explanations for how the 30% rule in AI can be interpreted, others might interpret it to mean that 30% of the company's investments should go toward ensuring quality and governance over data, ensuring workers stay productive, while also keeping the risks of using AI to a manageable level. Given that Gen Z can't stand AI, guidelines like this could help alleviate concerns about widespread AI adoption, especially in the workplace.