AI companies threaten democracy by pillaging journalism, but the press can fight back by exploiting their dependence on quality content
Source: https://www.theguardian.com/profile/carine-fouteau. "AI threatens democracy. But we are not powerless: here’s how to fight back | Carine Fouteau | The Guardian." October 16, 2026. www.theguardian.com
The Gist
The author argues that big AI companies are taking journalists' work for free and drawing readers and money away from news outlets, which endangers the free press that democracy relies on. But because AI products need good journalism to work well, news organisations have leverage and should fight back together through lawsuits, lobbying and public persuasion.
Conclusion
Journalists and the wider public-interest community should collectively pressure AI companies through legal action, lobbying and public persuasion to respect press and public rights, because AI firms are destroying the journalism that sustains democracy while depending on that same journalism.
Premises
- AI companies use news outlets' work without authorisation or compensation, and their summaries divert audiences and revenue from publishers, with profit rather than the public interest as their mission.
- Quality journalism differs fundamentally from AI output: it involves human fieldwork, verification, ethical standards, context and accountability, whereas generative AI only regurgitates and condenses existing material, often producing falsehoods, and its makers avoid legal liability for its reliability.
- If this appropriation continues unchecked, public-interest journalism will be destroyed, and with it the democratic public sphere built on reliable, pluralistic information, leaving a space governed by algorithms that amplify fake news and propaganda.
- AI companies depend on quality information: OpenAI has admitted that training leading models without copyrighted material is impossible, and a Microsoft executive has acknowledged that quality content improves output. By starving their sources, they risk degrading their own products into 'AI slop'.
- This dependence gives publishers leverage, which can be exercised through collective action, including litigation (such as the New York Times's suit and French proceedings over neighbouring rights), lobbying national and European lawmakers, building alliances with the research and knowledge community, and persuading the public of journalism's democratic value.
Assumptions
- Public-interest journalism is essential to a functioning democracy and has no adequate substitute.
- AI-generated summaries substantially substitute for reading original journalism, causing material harm to publishers' audiences and revenue.
- Using copyrighted news content for training and summarisation is unlawful or illegitimate and not covered by fair use or similar exceptions.
- AI companies need ongoing access to fresh, high-quality journalism, not just past archives, to remain viable.
- Legal and political pressure will be effective enough to change AI companies' behaviour before journalism is irreparably damaged.
- The interests of AI companies are sufficiently unified and hostile to the public interest that cooperation or licensing is not a sufficient remedy.