Fix AI’s data theft problem with onchain attribution

Ram Kumar, core contributor at OpenLedger, argues that people whose online posts, blogs and forum replies trained AI models should be paid through onchain attribution systems, a model he calls Payable AI. He proposes tagging each training item with a digital receipt and sending a small payment to the original creator each time the model is used, similar to music royalties and credited open-source code. He warns that without such tracking, AI could become a black market of stolen insights, and that intelligence could end up controlled by a few companies such as OpenAI, Meta and Google.