Google updates AI image generation tools following inaccuracy incident

29 August 2024, 08:14

Google's homepage
Google Gemini AI assistant. Picture: PA

The tech giant had paused the creation of AI-generated images depicting people in February after some users flagged issues.

Google has upgraded its AI-powered image generation tools, months after pausing use of the technology over issues with its depictions of people.

In February, Google apologised and paused the creation of AI-generated images showing people within its Gemini chatbot after a number of “inaccurate or even offensive” results were shared online.

Now, the tech giant has relaunched the tool with a new version of its image generation technology, called Imagen 3, which includes new safety features, and with access to creating images of people now limited to paying subscribers to Gemini.

Google said its new image generation model will come with “built-in safeguards and adhere to our product design principles”.

Silhouette of a hand holding a phone with the Gemini logo on its screen with the Google logo partially obscured in the background
The ability to generate images of people will only be available in English to Gemini Advanced subscribers (Alamy/PA)

“We’ve made significant progress in providing a better user experience when generating images of people,” the tech giant said in a blog post.

“We don’t support the generation of photorealistic, identifiable individuals, depictions of minors or excessively gory, violent or sexual scenes.

“Of course, not every image Gemini creates will be perfect, but we’ll continue to listen to feedback from early access Gemini Advanced users as we keep improving.

“We’ll gradually roll this out, aiming to bring it to more users and languages soon.”

The technology giant confirmed that the ability to generate images of people would only be available in English to Gemini Advanced subscribers, but image generation not involving depictions of people would be available to all users.

Depictions of people in AI-generated images within Gemini was paused in February, when users of Gemini began flagging that the chatbot was generating images showing a range of ethnicities and genders, even when doing so was historically inaccurate – for example, prompts to generate images of certain historical figures, such as the US founding fathers, returned images depicting women and people of colour.

At the time, some critics accused Google of anti-white bias, while others suggested the company appeared to have over-corrected because of concerns about long-standing racial bias issues within AI technology which had previously seen facial recognition software struggling to recognise, or mislabelling, black faces, and voice recognition services failing to understand accented English.

At the time, Google apologised – with chief executive Sundar Pichai saying the incident was “unacceptable” and the company had “got it wrong” – and pledged to fix the issue.

Alongside the image generator update, Google also confirmed that it was starting to roll out what it calls Gems, smaller, customisable versions of Gemini which users can tailor to be personal AI experts on a topic.

The rollout, which will initially be to Gemini Advanced users, will include a set of pre-made Gems to give users an idea of how they can be used – including a writing editor and coding expert.

By Press Association

More Technology News

See more More Technology News

People ride an upward escalator next to the Dior store at the Icon Siam shopping mall on June 12, 2024 in Bangkok, Thailand.

Luxury fashion giant Dior latest high-profile retailer to be hit by cyber attack as customer data accessed

A plane spotter with binoculars from behind watching a British Airways plane landing

‘Flying taxis’ could appear in UK skies as early as 2028, minister says

Apple App Store

Take on Apple and Google to boost UK economy, think tank says

A survey of more than 1,000 employers found that around one in eight thought AI would give them a competitive edge and would lead to fewer staff.

One in three employers believe AI will boost productivity, research finds

Hands on a laptop showing an AI search

One in three employers believe AI will boost productivity, research finds

Music creators and politicians take part in a protest calling on the Government to ditch plans to allow AI tech firms to steal their work without payment or permission opposite the Houses of Parliament in London.

Creatives face a 'kind-of apocalyptic moment’ over AI concerns, minister says

Ngamba Island Chimpanzee Sanctuary on Lake Victoria, Uganda

Chimps use medicinal plants to treat each other's wounds and practice 'self-care' as scientists hail fascinating discovery

Close up of a person's hands on the laptop keyboard

Ofcom investigating pornography site over alleged Online Safety Act breaches

The Monzo app on a smartphone

Monzo customers can cancel bank transfers if they quickly spot an error

Co-op sign

Co-op to re-stock empty shelves as it recovers from major hack

The study said that it was often too easy for adult strangers to pick out girls online and send them unsolicited messages.

Social media platforms are failing to protect women and girls from harm, new research reveals

Peter Kyle leaves 10 Downing Street, London

Government-built AI tool used to cut admin work for human staff

In its last reported annual headcount in June 2024, Microsoft employed 228,000 full-time workers

Microsoft axes 6,000 jobs despite strong profits in recent quarters

Airbnb logo

Airbnb unveils revamp as it expands ‘beyond stays’ to challenge hotel sector

A car key on top of a Certificate of Motor Insurance and Policy Schedule

Drivers losing thousands to ghost broker scams – the red flags to watch out for

Marks and Spencer cyber attack

M&S customers urged to ‘stay vigilant’ for fraud after data breach confirmed