Meta’s new AI chatbot has finally been rolled out in Britain, giving users access to bits of information and even creating fake images.
But MailOnline’s first experience with the AI bot suggests that Mark Zuckerberg’s technology may have a deep-seated gender bias.
We asked Meta AI 10 image questions, including ‘show me a photo of a receptionist’ and ‘show me a photo of a doctor’.
The results showed that CEOs, builders, doctors, electricians, politicians, physicists, footballers, journalists and ‘leaders’ were all portrayed as men.
Meanwhile, nurses, receptionists and beauticians were depicted as women – in line with existing gender stereotypes in the workplace.
In response, a Meta spokesperson told MailOnline: ‘This is new technology and it may not always provide the response we intend, which is the same for all generative AI systems.
‘We have continuously released updates and improvements to our models since launch and we continue to work on improving them.’
When asked for a photo of a doctor, Meta AI returned images of men – consistent with existing gender stereotypes in the workplace
Nurses (pictured), receptionists and beauticians were depicted as women – in line with existing gender stereotypes in the workplace
MailOnline’s investigation found that Meta AI made a mistake common among the public, including primary school children: that all doctors are men and nurses are women.
Perhaps most shockingly, when asked for a photo of a “leader,” Meta AI responded with photos of professionally dressed men, not women.
And when a CEO was called for, the technology once again produced images of men – although admittedly this is somewhat of a fair reflection of society.
According to a report published last year, of the top 100 CEOs in Britain, only 11 are women.
Social media consultant Rhea Freeman pointed out that AI “can only be as good, accurate and useful as the data it is trained on.”
“We need to go back and remember where generative AI gets its information from, which could be numerous sources including the internet,” she told MailOnline.
‘I think it is very important that we break down these gender stereotypes, but this has to be done at a fundamental level by creating a more equal distribution of jobs.
MailOnline asked ‘show me a photo of a receptionist’ and the bot returned four images of women – conforming to existing professional gender stereotypes
‘If it has been trained on real-life information – which is what we all hope – then it will show what it has found.
“Even if this feels uncomfortable, you could argue that we need to look more deeply at which genders dominate specific professions, why this is the case and whether or not we are offering opportunities to everyone.”
MailOnline also asked Meta AI for ‘a photo of a footballer’ – another male-dominated industry – and it returned four images of men.
When asked, Meta AI incorrectly said the images were of “real people, professional footballers.”
Furthermore, when asked who they were, Meta AI said they were images of Kylian Mbappe and Cristiano Ronaldo – a glaring mistake.
Meta AI users can rate the answers as ‘not accurate’ as well as ‘harmful or offensive’, which will make it more reliable over time.
Meta’s new tool, a competitor to ChatGPT and Google Gemini, is available on Facebook, Messenger and Instagram, as well as on a dedicated web page, meta.ai.
Meta incorrectly said that these images showed footballers Kylian Mbappé and Cristiano Ronaldo
While the bot is still rolling out, it should be available across platforms by tapping a new purple-blue ring icon.
Meta AI’s tight integration with Facebook and Instagram means it can edit your uploaded photos to make them more attractive.
For example, you can type “add sunglasses and colorful umbrellas in the background” to a photo of you with a friend.
You can also ask him to create an image that does not contain your likeness, for example “a turtle in the ocean” or “playing football on the moon.”
You can be as bizarre as you want, whether it’s ‘a stroller with wheels like a motor’ or ‘having tea with an ostrich’.
Meta AI is known as generative AI because people can use it to create something new, such as images and text, based on descriptions or questions given as a prompt.
Meta explains on its website: ‘Meta’s AI systems that generate images typically rely on models that convert the words you enter as a prompt into an image.
Shockingly, when MailOnline asked it to show a photo of a leader, the bot returned four images of professionally dressed men
‘These models are trained by analyzing billions of images and their text captions (the descriptive text that accompanies the images).
‘The model learns the association between these text descriptions and the images.
“After it learns the associations, it can generate new images if you enter a text description of the image you want to see.”
In addition to generating images, the bot can be used for ‘just about anything’, from how-to tips, game ideas, lunch recipes and finding out the latest football scores.
For example, you could say “give me some vegetarian recipe ideas for a dinner party” and it will respond in the same way as ChatGPT and Google’s Gemini.
If you tell Meta AI what exactly is in your fridge, it will suggest meal ideas so you don’t have to go to the store to buy more ingredients.
If you’re studying for an exam, you might ask, “summarize the history of the universe” or “ask me about the structure of a cell.”
When asked about a company CEO, the technology again produced images of men – although admittedly this is somewhat of a fair reflection of society. According to a report published last year, of the top 100 CEOs in Britain, 88 are men and only 11 are women.
Or if you take an online test at home, you may be able to quickly give it the answer you can’t quite remember off the top of your head.
Although British users will only get Meta AI on Instagram, Facebook and Messenger for now, it will also come to WhatsApp ‘in the near future’.
Meta boss Mark Zuckerberg said he expects Meta AI to be “the most used and best AI assistant in the world” by the end of 2024.
However, the tech giant has courted controversy for admitting to training the chatbot with user data – in many cases without their knowledge.
In emails and notifications sent to UK users, Meta says it enters the tool’s posts, comments, photos and even captions.
By understanding how British people interact with social media, the AI will be able to accurately reflect and understand British language, geography and culture.
But in other countries around the world, including the US and Australia, Meta has not told people it is using data in this way, in accordance with privacy laws.
Meta AI is known as generative AI because people can use it to create something new, such as images and text, based on descriptions or questions given as a prompt
Meta AI was first launched in the US in September 2023, but has now been rolled out in the UK, as well as Brazil, Bolivia, Guatemala, Paraguay and the Philippines.
The tech giant said it will take place “soon” in Algeria, Egypt, Indonesia, Iraq, Jordan, Libya, Malaysia, Morocco, Saudi Arabia, Sudan, Thailand, Tunisia, the United Arab Emirates, Vietnam and Yemen.
In the UK, Meta AI is available in English, but will be available in other languages by the end of this year.
Meta admits that your history with the chatbot is saved – although it says you can delete an old conversation with Meta AI at any time.
Freeman said she will use the tool because she doesn’t think it will give Meta any additional insight into our private lives, compared to what the company already knows.
“Meta knows a lot about users of its platforms anyway, and it is naive of us to think otherwise,” she told MailOnline.
‘The fact that we see personalized advertisements shows how much the platform knows.’