NEW YORK (AP) 鈥 OpenAI on Monday said it plans to halt the use of one of its ChatGPT voices that 鈥淗er鈥 actor Scarlett Johansson says sounds 鈥渆erily similar" to her own.
In a on the social media platform X, OpenAI said it is 鈥渨orking to pause鈥 Sky 鈥 the name of one of five voices that ChatGPT users can chose to speak with. The company said it had 鈥渉eard questions鈥 about how it selects the lifelike audio options available for its flagship artificial intelligence chatbot, particularly Sky, and wanted to address them.
Among those raising questions was Johansson, who famously voiced a fictional, and at the time futuristic, AI assistant in the 2013 film
Johansson issued a statement saying that OpenAI CEO Sam Altman had approached her in September asking her if she would lend her voice to the system, saying he felt it would be 鈥渃omforting to people鈥 not at ease with the technology. She said she declined the offer.
鈥淲hen I heard the released demo, I was shocked, angered and in disbelief that Mr. Altman would pursue a voice that sounded so eerily similar to mine that my closest friends and news outlets could not tell the difference,鈥 Johansson said.
She said OpenAI 鈥渞eluctantly鈥 agreed to take down the Sky voice after she hired lawyers who wrote Altman letters asking about the process by which the company came up with the voice.
OpenAI had moved to debunk the internet's theories about Johansson in a blog post accompanying its earlier announcement aimed at detailing how ChatGPT's voices were chosen. The company that it believed AI voices 鈥渟hould not deliberately mimic a celebrity鈥檚 distinctive voice鈥 and that the voice of Sky belongs to a 鈥渄ifferent professional actress." But it added that it could not share the name of that professional for privacy reasons.
In a statement sent to The Associated Press following Johansson's response late Monday, Altman said that OpenAI cast the voice actor behind Sky 鈥渂efore any outreach" to Johansson.
鈥淭he voice of Sky is not Scarlett Johansson鈥檚, and it was never intended to resemble hers,鈥 Altman said. 鈥淥ut of respect for Ms. Johansson, we have paused using Sky鈥檚 voice in our products. We are sorry to Ms. Johansson that we didn鈥檛 communicate better.鈥
San Francisco-based OpenAI first rolled out voice capabilities for ChatGPT, which included the five different voices, in September, allowing users to engage in back-to-forth conversation with the AI assistant. 鈥淰oice Mode鈥 was originally just available to paid subscribers, but in November, OpenAI that the feature would become free for all users with the mobile app.
And ChatGPT's interactions are becoming more and more sophisticated. Last week, OpenAI said the can mimic human cadences in its verbal responses and can even try to detect people鈥檚 moods.
OpenAI says the newest model, dubbed GPT-4o, works faster than previous versions and can reason across text, audio and video in real time. In a demonstration during OpenAI's May 13 announcement, the AI bot chatted in real time, adding emotion 鈥 specifically 鈥渕ore drama鈥 鈥 to its voice as requested. It also took a stab at extrapolating a person鈥檚 emotional state by looking at a selfie video of their face, aided in language translations, step-by-step math problems and more.
GPT-4o, short for 鈥渙mni,鈥 isn't widely available yet. It will progressively make its way to select users in the coming weeks and months. The model's text and image capabilities have already begun rolling out, and is set to reach even some of those that use ChatGPT's free tier 鈥 but the new voice mode will just be available for paid subscribers of ChatGPT Plus.
While most have yet to get their hands on these newly announced features, the capabilities have conjured up even more comparisons to the Spike Jonze's dystopian romance 鈥淗er," which follows an introverted man (Joaquin Phoenix) who falls in love with an AI-operating system (Johansson), leading to many complications.
Altman appeared to tap into this, too 鈥 simply the word 鈥渉er鈥 on the social media platform X the day of GPT-4o's unveiling.
Many reacting to the model's demos last week also found some of the interactions struck a strangely flirtatious tone. In one posted by OpenAI, a female-voiced ChatGPT compliments a company employee on 鈥渞ocking an OpenAI hoodie,鈥 for example, and in another the chatbot 鈥渙h stop it, you're making me blush鈥 after being told that it's amazing.
That鈥檚 sparked some conversation on the gendered ways critics say tech companies have long used to develop and engage voice assistants 鈥 dating back far before the latest wave of generative AI advanced the capabilities of AI chatbots. In 2019, the United Nations' culture and science organization built into default female-voiced assistants (like Apple's Siri to Amazon's Alexa), even when confronted with sexist insults and harassment.
鈥淭his is clearly programmed to feed dudes鈥 egos,鈥 The Daily Show senior correspondent Desi Lydic said of GPT-4o in a segment last week. 鈥淵ou can really tell that a man built this tech.鈥
___
The Associated Press and OpenAI have that allows OpenAI access to part of AP鈥檚 text archives.