ChatGPT suspends its new ‘flirty’ AI voice over fears it sounds too much like Scarlett Johansson
ChatGPT’s creator OpenAI has stopped using one of its AI-generated voices after some users said it had a flirty tone that sounded too much like Hollywood A-Lister Scarlett Johansson.
In a message to
The company said it had “heard questions” about how it selects the lifelike audio options available for its flagship artificial intelligence chatbot, specifically Sky, and wanted to answer them.
OpenAI also quickly debunked internet theories about Johansson in an accompanying blog post detailing how ChatGPT’s votes were chosen.
“We believe that AI voices should not intentionally mimic a celebrity’s signature voice – Sky’s voice is not an imitation of Scarlett Johansson, but belongs to another professional actress using her own natural speaking voice,” the company wrote .
Some users said the tone had a flirty tone that sounded too much like Hollywood A-Lister Scarlett Johansson (pictured)
OpenAI said it ‘worked to pause Sky’ – the name of one of five voices ChatGPT users can choose to talk to (file image)
Johansson starred in the 2014 Spike Jonze film “Her,” in which she voiced an AI that Joaquin Phoenix’s character falls in love with.
ChatGPT users found that some interactions with Sky were bizarre flirtations. In a video posted by OpenAI, a female ChatGPT voice compliments a company employee for “wearing an OpenAI hoodie.”
In another, the chatbot says ‘oh stop it, you’re making me blush’ after being told it’s great.
“This is clearly programmed to feed guys’ egos,” The Daily Show senior correspondent Desi Lydic said in a segment last week. “You can really tell that a man built this technology.”
OpenAI first rolled out voice capabilities for ChatGPT in September, including the five different voices, allowing users to have back-to-for conversations with the AI assistant.
‘Voice Mode’ was originally only available to paying subscribers, but in November OpenAI announced that the feature would become free to all users with the mobile app.
Johansson starred in the 2014 Spike Jonze film ‘Her’, in which she voiced an AI that Joaquin Phoenix’s character falls in love with (pictured)
OpenAI CEO Sam Altman appeared to capitalize on the Her comparison by simply posting the word “her” on social media platform X on the day of GPT-4o’s unveiling (File Image)
And ChatGPT’s interactions are becoming increasingly sophisticated. Last week, OpenAI said the latest update to its generative AI model can mimic human cadences in its verbal responses and even attempt to detect people’s moods.
OpenAI says the latest model, called GPT-4o, runs faster than previous versions and can reason through text, audio and video in real time.
During a demonstration during OpenAI’s May 13 announcement, the AI bot chatted in real time, adding emotion — specifically “more drama” — to its voice as requested.
It also took effort to extrapolate a person’s emotional state by looking at a selfie video of their face, aided by language translations, step-by-step math problems and more.
GPT-4o, short for ‘omni’, is not yet generally available. It will gradually make its way to user selection over the coming weeks and months. The model’s text and image capabilities have already been rolled out and will even reach some of those using the free version of ChatGPT, but the new voice mode will only be available to paid subscribers of ChatGPT Plus.
While most haven’t gotten their hands on these newly announced features yet, its capabilities have sparked even more comparisons to Spike Jonze’s dystopian romance “Her.”
Sam Altman, CEO of OpenAI, also seemed to take advantage of this by simply posting the word “her” on the social media platform X on the day of GPT-4o’s unveiling.