Meta is reportedly engaged in negotiations with prominent Hollywood actors to incorporate their voices into its AI systems. The goal is to unveil these features during the Meta Connect conference scheduled for September. According to Bloomberg, the company is in talks with notable celebrities such as Judi Dench, Keegan-Michael Key, and Awkwafina, aiming to finalize multi-million dollar contracts.
Celebrity Voices in AI
This new development involves the Meta AI digital assistant. Meta is deploying this project to enhance user experience across its platforms. The New York Times reports that the firm is collaborating with top talent agencies in Hollywood to secure these celebrity voices. The agreements are expected to be short-term but renewable, covering platforms like Facebook, Instagram, and Meta's Ray-Ban smart glasses.
This effort comes after Meta discontinued its previous celebrity AI chatbots, which were presented at the last Connect event. The company has made substantial investments in AI technologies, such as the open-source AI model LLaMA 3.1. Though the concept of using recognizable voices in AI isn't entirely new—seen in applications like Waze and Amazon's Alexa—these earlier implementations often felt more gimmicky than essential.
Potential Impact and Challenges
The advancement of generative AI has made it increasingly feasible to mimic human voices with a high degree of accuracy. Nevertheless, ethical and legal issues persist. Even so, Big Tech companies continue to explore this section of AI. For example, Google is ready to offer its own celebrity chatbots, an ElevenLabs is creepily using the voices of deceased stars – with permission form their estates – to narrate content.
While these tools seem harmless, the dangers of deepfake content driven by AI is a real one. We have already seen OpenAI come under fire for its Voice feature in ChatGPT. Unveiled during OpenAI's Spring Update event in May 2024, the feature initially used a voice named Sky, which resembled Scarlett Johansson. Johansson objected, revealing she had declined to lend her voice for ChatGPT. OpenAI quickly canned the feature, made some tweaks, and is now rolling it out again.
