Can paying artists fix the AI copyright mess?

The Fender guitar company recently faced backlash after its CEO, Bud Cole, suggested that playing cover songs and collaborating with bandmates is a form of analog AI. He argued that AI could help musicians learn to write songs, effectively acting as an assistant. Meanwhile, a new AI video startup called Pippa is trying a different approach to bridge the divide. They offer a revenue-sharing system where artists receive small payments whenever a user generates content in their specific style. Pippa is currently attempting to sign up artists to train their AI models on original work, hoping to create a more ethical version of generative video tools.

To understand why this is so controversial, you have to look at how these AI models are actually built. Most AI models are created through a process called training, where a computer analyzes millions of examples—like songs, paintings, or videos—to recognize patterns. Think of it like a student reading an entire library to learn how to write. Once the model learns these patterns, it can generate new output based on a prompt. The problem is that many of these models were originally trained by scraping the internet for data, often without the consent or compensation of the original creators. When an AI produces a new video in the style of a specific artist, it is using those learned patterns. While Pippa is trying to pay artists, their systems still rely on existing, massive models that were likely built on that same pool of uncompensated, scraped data.

The core disagreement is whether human learning and AI training are the same thing. Critics argue that human musicians learn through lived experience, emotional choices, and physical practice, whereas AI functions by identifying statistical correlations in massive datasets. Relying on AI to shortcut the songwriting process might lead to what experts call deskilling, where the act of creating art is reduced to prompt-based output rather than the hard work of practice and repetition. As we look ahead, the big question is whether financial compensation is enough to satisfy creators, or if the very mechanism of using AI to mimic human style remains fundamentally at odds with the spirit of the arts. We are watching a slow, painful negotiation over who gets credit and payment in a world where machines can imitate the human touch.