AI, Copyright, and the Ethics of Consent

The technology may be new. But the imbalance of power is not.

In March 2025, OpenAI, later joined by Google, urged the U.S. government to clarify copyright rules to allow AI companies to train their models on protected content under the principle of “fair use.” Their framing is strategic: they claim it’s about maintaining U.S. leadership in artificial intelligence, especially compared to countries with more permissive data environments.

But underneath the policy argument lies a deeper ethical question. If the content powering these models is created by people — artists, writers, journalists, photographers, educators — shouldn’t they have a say in how it is used?

Whose stories feed the machine?

From image libraries to language archives, the raw material that trains AI systems is often gathered with little transparency or accountability. Content is scraped at scale. Consent is assumed or ignored. Attribution is rare. Compensation is almost never on the table.

This extractive logic isn’t new. It mirrors colonial patterns of knowledge appropriation, where the insights, expressions, and experiences of others are repurposed to serve the goals of those in control.

At Fairpicture, we recognize these patterns because they also show up in visual storytelling. For decades, narratives in the development and humanitarian sectors have been shaped by outsiders, filtered through the lenses of power, privilege, and distance. Communities have been portrayed as passive, voiceless, or defined by lack, all in the service of someone else’s story.

It’s part of what drives our commitment to more thoughtful storytelling.

We believe ethical storytelling must hold, especially now.

That means:

• Informed consent must be a baseline, not a bonus. No individual, artist, or community should find their likeness or life embedded in a dataset they never agreed to be part of.
  
  
• Attribution and compensation are not optional. If creative work fuels billion-dollar models, its creators deserve to be recognised and paid.
  
  
• Cultural accountability matters. Without it, AI risks amplifying the very dynamics that global justice movements seek to dismantle, from racial bias to epistemic injustice.
  

Visual justice in an AI world

Fairpicture was founded on the belief that content creation should be ethical, equitable, and inclusive, not extractive. We work with local creators, embedded in their communities, to co-create visual stories with dignity, context, and care. That model may be slower, yes. But it is also richer, more just, and more sustainable.

In a world where AI models can generate endless images and text at lightning speed, we’re reminded that the best stories are not the fastest. They are the ones built through relationship, trust, and respect.

If AI is to be part of our storytelling future, it must be built on different terms. Terms that honor human agency, not erase it.