Is it legal to train AI on copyrighted books? It’s complicated
24 August 2026 · filed under 8dafa127b694
TechCrunch reports on the unresolved legal status of training AI systems on copyrighted books. According to the outlet, most published authors have contributed, without their knowledge or consent, to the development of AI tools that threaten to undermine their livelihoods.
The piece frames the central question plainly: given that contribution, is the practice legal. TechCrunch characterizes the answer as complicated, per its own headline and summary, without resolving the question one way or another.
The report does not specify which authors, publishers, or AI companies are involved, nor does it detail how the contribution occurred beyond the general description offered. It stops short of naming specific datasets, systems, or mechanisms by which authors’ work became part of AI development.
TechCrunch’s framing ties the matter directly to authors’ livelihoods, describing the AI tools in question as posing a threat on that front. The outlet does not elaborate further on the nature or extent of that threat within the material provided.
The Specola notes this as a report establishing a question rather than an answer. TechCrunch’s own characterization, “It’s complicated,” stands as the outlet’s summary judgment on the state of the issue as of publication, August 23, 2026.
No further particulars, rulings, or elaborations beyond this framing appear in the source material reviewed for this bulletin.