The capability to mimic someone’s voice so precisely that it becomes indistinguishable from the original has significant implications. As reported by BBC News, the faked audio involving Mayor Sadiq Khan did not amount to a crime according to the Met Police, yet this instance exposes how deepfakes can be used nefariously by opposing political parties or other malicious actors. In parallel, VentureBeat discloses the FTC’s concerns over voice cloning AI—highlighting its potential for consumer fraud and misuse. With the absence of comprehensive protective measures, anyone’s vocal identity could easily become ammunition for deception and manipulation.
Universal Music Group's response to an unauthorized AI-generated song featuring Drake and The Weeknd underlines the intellectual property debate intensifying around AI. As documented by Document Journal, the company took swift action to block the use of copyrighted material for creating AI-generated content without consent. This situation exemplifies how current copyright laws fall short when confronting the novel challenges posed by generative AI technologies.
Innovations in AI have outpaced the legal frameworks intended to protect creators’ rights. In a previous blog I discuss the rise of decentralized identity management through blockchain as a transformative solution. However, until legislators catch up, there exists a grey area where personalities can be exploited without repercussion, highlighting the urgent necessity for likeness detection systems coupled with fair compensation mechanisms. YouTube has an implementation of this called Content ID - however it is only available to a few select creators.
The evolution of voice synthesis technology necessitates robust regulation to ensure individual protections. Echoing the sentiment from "Sneakers" - "You won't know who you trust." - as the notion of trust becomes a rare commodity when voices and likeness can be replicated flawlessly by machines. It is clear as day that regulation which affords individual protections is imperative. Adequate guardrails in the form of likeness detection tools and decentralized remuneration systems must be established rapidly to mitigate the risks associated with generative AI. Without these safeguards, we risk entering a world where neither artistic integrity nor personal security can be assured.
As an aside - as there is no ownership guideline for generative works, the image for this article has been directly lifted from VentureBeat.
Copyright (c) Vohra Pte Ltd 2020 - 2024
Vohra Pte Ltd is registered in Singapore
Examining the UK's Reluctance to Regulate AI and Its Potential Implications