You’re witnessing journalism’s latest existential crisis unfold in real time, and New York Times publisher A.G. Sulzberger isn’t pulling punches. He’s calling AI companies’ data practices “brazen theft of intellectual property at an unprecedented scale,” accusing tech giants of strip-mining news websites without permission then repackaging the content as their own. The Times has already burned $20 million fighting OpenAI, Microsoft, and Perplexity in court—but Sulzberger insists they had no choice.
The stakes extend far beyond journalism’s financial woes. Sulzberger warns that AI’s content raid threatens a global creative economy worth $12 trillion annually, employing over 50 million people across publishing, film, music, and academia. Think Netflix writers, Spotify artists, and the photographers whose work trains image generators. When AI platforms become your primary news source, those creators lose both audience and revenue that funds their next project.
Here’s where it gets weird: media companies are simultaneously suing and partnering with the same AI firms. News Corp signed licensing deals with OpenAI and Meta while Dow Jones and the New York Post sue Perplexity for alleged copyright theft. CNN’s parent company Warner Bros. Discovery also filed against Perplexity, proving even broadcast giants see AI summarization as an existential threat. It’s like arguing with your landlord while paying rent—desperate times demand contradictory strategies.
The math looks brutal for human creators. CISAC projects that by 2028, music creators could lose 24% of their revenue and audiovisual creators 21%, totaling €22 billion in potential annual losses as AI systems both use their work without payment and replace demand for originals. Meanwhile, AI-generated content revenues in those sectors could hit €9 billion annually, mostly from unlicensed material.
Whether you get tomorrow’s news from human journalists or AI summaries depends on how courts rule on fair use, whether sustainable licensing models emerge, and if platforms design products to send traffic back to original sources. The alternative isn’t just fewer reporters—it’s a future where algorithms trained on yesterday’s creativity become the only voice left standing.
