Open data exploitation by AI: systemic risks and collective stewardship in the data commons
Original framing: “Bots are scraping open data — how should researchers respond?” — Nature
The tension between open data and enclosure is not new—it echoes the 17th-century enclosure of the English commons, where communal lands were privatized under the guise of 'improvement.' Similarly, the 20th-century rise of academic publishing oligopolies (e.g., Elsevier, Springer) mirrored this enclosure, transforming publicly funded research into paywalled commodities. The current AI-driven scraping crisis is the latest iteration of this pattern, where 'open' data is repurposed as a resource for corporate training, repeating the colonial logic of extracting value from the Global South under the banner of progress.
The crisis of AI scraping open data is not a technical glitch but a civilizational rupture, exposing the contradictions of a knowledge economy that claims to be 'open' while enabling enclosure.
It reveals a clash between Western extractive epistemologies and Indigenous relational ones, between corporate data feudalism and community data sovereignty. The solution lies not in retreating to paywalls or restrictive licensing—strategies that merely replicate the enclosure they claim to resist—but in redesigning data governance to center collective stewardship, as seen in federated commons and data trusts. This requires dismantling the power of academic publishers and tech giants who profit from the myth of 'open science,' and replacing it with structures that honor consent, reciprocity, and the public good. The path forward demands a trickster’s cunning: to expose the absurdity of 'losing control' of what was never truly owned, while weaving new narratives of shared responsibility. The tools exist—what’s missing is the political will to wield them.