Undermining creative economies
LMs may generate content that is not strictly in violation of copyright but harms artists by capital- ising on their ideas, in ways that would be time-intensive or costly to do using human labour. This may undermine the profitability of creative or innovative work. If LMs can be used to generate content that serves as a credible substitute for a particular example of hu- man creativity - otherwise protected by copyright - this potentially allows such work to be replaced without the author’s copyright being infringed, analogous to ”patent-busting” [158] ... These risks are distinct from copyright infringement concerns based on the LM reproducing verbatim copyrighted material that is present in the training data [188].
ENTITY
2 - AI
INTENT
2 - Unintentional
TIMING
2 - Post-deployment
Risk ID
mit229
Domain lineage
6. Socioeconomic and Environmental
6.3 > Economic and cultural devaluation of human effort
Mitigation strategy
1. Establish robust and enforceable fair remuneration mechanisms. Mandate the redistribution of revenue generated by AI systems—especially when their training data incorporates human-created work—back to the original creative sources through new contractual standards, licensing requirements, or dedicated compensation funds. 2. Mandate the implementation of digital provenance and transparency tools. Require AI developers and deployers to embed technical safeguards, such as watermarking or metadata, that reliably and conspicuously distinguish AI-generated content from human-made originality, thereby preserving the commercial value and visibility of human effort. 3. Strengthen intellectual property frameworks to protect creative identity and style. Enact specific legislation to require explicit, informed consent and fair compensation for the use of an artist's distinctive style, voice, or likeness in AI model training and content generation, mitigating the risk of unauthorized emulation and substitution.
ADDITIONAL EVIDENCE
GPT-2 has been used to generate short stories in the style of Neil Gaiman and Terry Pratchett [178], and poems in the style of Robert Frost and Maya Angelou [81], suggesting that emulation of artist’s styles is possible (see also the VersebyVerse [184] tool) [77].