Valley must respect (Beverly) Hills

economictimes.indiatimes.com:

Tech companies are wading into copyright issues not only over the content they use to train AI but also over how it interacts with humans. Hollywood star Scarlett Johansson has raised the issue of informed consent and fair compensation after OpenAI launched a chatbot with a voice apparently similar to hers. OpenAI claims it has used the voice of another unnamed artiste for its Sky chatbot.

But it has paused its rollout after Johansson's complaint. The AI company had initially approached her with a contract for use of her voice but she had turned it down — twice. The film industry has reason to be suspicious of AI and has kitted itself up with legal protection against improper use of creatives by the tech industry, particularly after a rash of deepfakes involving some of its biggest celebrities.

Creators of AI will have to be especially careful over copyright, given the scope for bias inherent in the technology.

Bias enters through human input. But AI can amplify it. Creative people lending names, faces or voices to such endeavour would be justified in seeking protection to their livelihood apart from regular compensation.

This would also apply to content AI trains upon, such as the work of writers, directors and musicians.

Changes to IPR protection are required in countries where such systems are robust. Elsewhere, like in India, protection needs to be introduced and executed. The human interface of AI will multiply as it spreads across cultures and languages.

The need for creative inputs will rise accordingly. The scope for consent and compensation infringement will grow as AI disperses through economic systems. Lesser brands than OpenAI and less famous celebrities than Johansson must have strong regulation to ensure