Introduction: In today’s digital age, the concept of property, particularly intellectual property, is rapidly evolving. A recent study co-authored by Bart Wilson and myself explores this shift, revealing insights into how people perceive ownership and sharing in virtual environments.
I have a new working paper with Bart Wilson titled: “You Wouldn’t Steal a Car: Moral Intuition for Intellectual Property”.
The title of this post, “everyone take copies,” originates from discussions among participants in an experiment at our lab that serves as the foundation for our paper. This experiment investigates the dynamics of resource sharing and appropriation.
In this experiment, each participant controls a round avatar within a virtual space, as illustrated in the screenshot below.

The “seeds” in our experiment represent rivalrous resources, akin to physical goods, where ownership cannot be shared simultaneously. For example, if the Almond-colored player takes a seed from the Blue player, the Blue player loses that seed—similar to a car being stolen from its owner.
As a result, players often refer to this act of taking seeds as “stealing,” as indicated in the speech bubble in the accompanying picture. This observation aligns well with Bart Wilson’s prior research on the foundational aspects of physical property.
Our research also explores whether similar attitudes arise when it comes to non-rivalrous goods, which we designate as “discs.” Unlike rivalrous goods, multiple individuals can utilize non-rivalrous goods without diminishing their availability for others. Therefore, if one participant takes a disc, the original holder still retains access to their disc.
While participants are not obligated to use the chat feature, many choose to engage with each other actively, forming a community to share resources. Importantly, our findings indicate that subjects do not characterize the appropriation of digital goods (discs) as “stealing.”
In our paper, we note:
Participants frequently discuss discs, highlighting their understanding of the resource. They often articulate the positive-sum rationale behind zero-marginal-cost copying. For instance, farmer Almond states, “ok so disks cant be stolen so everyone take copies,” explicitly rejecting the notion of “theft” concerning discs.
Participants never admonish others for taking disc copies, yet frequently encourage them to refrain from taking seeds. The objection is directed at the removal of rivalrous goods, not the act of copying itself. As farmer Almond clarifies in noSeedPR2, “cuz if u give a disc u still keep it,” stressing that creators can replicate discs without incurring marginal costs.
We invite you to read the manuscript for an in-depth exploration of our experimental setup and the mechanisms of exchange. Our conclusion posits that, contrary to the intent of the “You Wouldn’t Steal a Car” campaign by the Motion Picture Association of America in the early 2000s, individuals do not instinctively view piracy as a criminal offense.
While people can articulate that digital piracy is illegal and take steps to prevent it, instilling a sense of guilt about piracy is challenging. This contrasts with the guilt felt when individuals believe they are harming someone directly.
These insights have notable implications for the structure of the contemporary information economy. One recent model is termed “the subscription economy,” where consumers increasingly opt for recurring payments for continued access to products and services (e.g., Netflix, Adobe software) rather than making one-time purchases. Gen Z, for example, has voiced concerns on platforms like TikTok about feeling trapped by recurring expenses and lacking true ownership of their digital assets.
In a recent talk show interview titled The Stream, I suggested that a possible motivation for companies transitioning to subscription models is their distrust in consumers regarding “ownership” of digital products. With the potential for individuals to share digital copies of music and software, companies often struggle to monetize these digital assets fully. The mantra “everyone take copies” reflects this challenge.
A unique aspect of our experimental design ensures that even when a disc is shared, the original creator rarely receives direct compensation, yet their authorship remains visible. For instance, a disc created by the Blue player is associated with their avatar’s color, making it clear who the contributor is. This approach allows the Blue player to witness their work being circulated.
The rise of large language models in recent information technologies also presents issues for idea creators, as many are often uncredited when their work influences responses generated by tools like ChatGPT. A recent settlement with Anthropic underscores this concern, as the company agreed to compensate for some written training material used in developing the AI model named Claude. How human creators are recognized (or not) for their contributions to AI systems will shape the future landscape of idea generation and ownership.
Conclusion: The research discussed reveals complex dynamics surrounding the perception of ownership and sharing in digital contexts. As society continues to navigate the intricacies of intellectual property, understanding these moral intuitions can inform the evolution of economic models and cultural attitudes towards digital goods.