Copyright expansion has failed to protect creative workers from wage erosion, and relying on intellectual property law alone to fight AI won't change that outcome. While many artists and creators are pushing for stricter copyright rules to prevent AI companies from using their work for training, tech policy experts argue this strategy misses the real problem: creative workers lack the bargaining leverage needed to negotiate fair compensation, regardless of how many legal restrictions exist.

Why Has Copyright Failed Creative Workers So Far?

Over the past 50 years, copyright law has expanded dramatically. Works are protected longer, restrictions cover more uses, penalties for infringement have grown steeper, and copyright now extends to more types of creative output. Yet during this same period, creative workers have grown poorer while their employers have grown richer.

The problem isn't the absence of legal rights. The problem is leverage. Copyright operates as a property right that creators can license and sell to media companies, publishers, studios, and other intermediaries. But this only works if creators have options and bargaining power. When the creative industries consolidate into just a handful of massive corporations, individual creators lose their negotiating position entirely.

Today's media landscape reflects this consolidation: five major publishers, four major studios, three major music labels, two dominant app stores, and one company controlling most ebooks and audiobooks. In this environment, even exceptional artists often have little choice but to accept whatever terms are offered. Copyright gives creators rights on paper, but not the power to enforce them in practice.

What's the Real Problem With an AI-Focused Copyright Strategy?

Many creators believe the solution is to craft new copyright laws that specifically ban scraping, analysis, and publication of copyrighted works when those activities are part of AI training. The theory sounds appealing: ban the bad uses while protecting good ones like search engines, digital archives, and academic research. But experts warn this approach faces serious practical obstacles.

According to tech policy analysis, distinguishing between "bad" AI training and "good" data analysis is far more difficult than it appears. Both processes involve the same technical steps: retrieving files from the internet, analyzing their contents, and publishing findings based on that analysis. Creating a statute that stops one but not the other may be nearly impossible without causing collateral damage to legitimate activities.

The collateral damage would be substantial. Search engines, the Oxford English Dictionary, the Internet Archive, and large-scale linguistic research all depend on the ability to collect and analyze copyrighted works without permission. Restricting these activities to stop AI training could undermine these socially beneficial services.

How Can Creative Workers Actually Protect Their Interests?

Rather than relying on copyright expansion, experts suggest creative workers should focus on labor rights and collective bargaining. The Writers Guild of America offers a real-world example: it's the only group of creative workers in history to comprehensively defeat AI in the workplace, and it did so through a strike, not through copyright litigation.

This approach reframes the problem. Instead of thinking of themselves as small businesses licensing intellectual property, creative workers should think of themselves as workers negotiating labor conditions. This shift opens different policy tools:

• Sectoral Bargaining: Collective agreements that cover all workers in an industry, giving workers negotiating power even when individual companies are large and dominant.
• Labor Protections: Workplace rules that prevent wage erosion and ensure fair compensation, similar to protections in other industries.
• Collective Action: Strikes and organized labor movements that have historically proven more effective than legal restrictions in securing worker rights.

The core insight is that property rights cannot substitute for labor rights. Giving creative workers more copyright without giving them more bargaining power is like giving a bullied student extra lunch money. The bullies simply become richer. But if you organize workers collectively and give them the ability to negotiate as a group, you change the power dynamic entirely.

This doesn't mean copyright is irrelevant. Rather, it means copyright alone is insufficient. Creative workers need both intellectual property protections and the labor rights necessary to enforce them. Without bargaining leverage, even the strongest copyright law will fail to protect livelihoods from AI-driven wage erosion.

What Does This Mean for the Current AI Copyright Debate?

The current legal battles over AI training and copyright are important, but they address a symptom rather than the underlying disease. Courts may rule that AI training violates copyright, or they may rule that it constitutes fair use. Either way, these decisions won't solve the fundamental problem: creative workers lack the bargaining power to negotiate fair compensation in a consolidated media landscape.

Some artists and creators understand this trade-off and are willing to accept collateral damage to other beneficial activities in exchange for stricter AI copyright rules. They've decided that imperfect protection is better than none. But policy experts argue this represents a losing strategy in the long term. The real path to protecting creative workers lies in building labor power, not in expanding intellectual property law.