Scarlett Johansson heats up ethical debate around generative AI

Actress claims that ChatGPT's talking robot imitated her voice, and OpenAI announces model suspension

Reputation Feed Writing

Scarlett Johansson, who voiced the character Samantha, a digital assistant in the film She, said she was shocked by the similarity between the robot’s voice and hers – Photo: Shutterstock

An update to ChatGPT has once again provoked reflections and heated discussions about ethical issues involving generative artificial intelligence. This time around actress Scarlett Johansson and one of five models of OpenAI's talking robots. According to the star, the model called Sky imitated his voice and the way he spoke.

The version has been available since September 2023, but debates gained momentum with the launch on May 13, from ChatGPT-4o, the new version of the bot, which makes robotic voices more natural and with more cadence, more similar to those of a human. The actress disclosed that she made a formal complaint to OpenAI, which announced the discontinuation of use of Sky.

What could be another success story for a company that has stood out against big tech in the generative AI segment, turned into a headache. After launch, many users who tested GPT-4o noticed a great similarity between Sky's voice and Scarlett's. The perception was heightened after social media was flooded with publications referring to the actress' role in the film. She (Her, Spike Jonze/2013), in which he voices Samantha, a digital voice assistant with whom the main character falls in love.

Through a note from her advisor, the actress said she was shocked by the adoption of the voice that imitated her mannerisms. She also reported that the OpenAI staff had contacted her, months before the launch, asking for licensing for the use of her voice in the application. However, she denied it due to personal reasons. According to Scarlett, the company would have made a new attempt two days before launching the update.

Após dias de polêmica nas redes, o presidente executivo da OpenAI, Sam Altman, também chegou a insinuar que a similaridade nas vozes de Sky e Scarlett era proposital, ao responder uma publicação no X (antigo Twitter) citando o nome do filme com a atriz.

When announcing the interruption, however, OpenAI defended itself by stating that the voice is not an imitation of Scarlett but, rather, natural and belonging to a professional actress. The company did not reveal the name of the voice's owner to protect her privacy.


SURVEY

How much do you consider the 2024 Elections to represent reputational risks for companies?

PUBLICATIONS RELATED

Avanço da Inteligência Artificial acende debate sobre riscos à reputação
Board
7 MIN.

Advancement of Artificial Intelligence sparks debate on reputational risks

Experts warn of the necessary care in the use of the technological tool by companies

Reputation Feed Writing
read more
Inteligência Artificial cria alertas para a reputação de empresas e líderes
Article
3 MIN.

Artificial Intelligence creates alerts for the reputation of companies and leaders

ChatGPT provokes reflection on the limits of its use and the risks for companies and their leaders

By Renata Afonso
read more
Tecnologia entre benefícios e riscos à reputação
Article
3 MIN.

Technology between benefits and risks to reputation

Companies must protect customer data and ensure the best possible use of solutions

By Leandro Balbinot
read more
Quem mais ganha reconhecimento com inovação
curated by ANK
2 MIN.

Who wins the most recognition with innovation

Fast Company ranking sees less and less difference between tech companies and the rest

Clovis Malta
read more
COMMENT

Fill out the form below to send your comment:

Check out the opinion of those who have already read this content:

(No comment)