Facebook, Instagram to label images generated by AI tools

Meta's president of global affairs says company will deploy technologies across Facebook, Instagram and Threads to detect and label such images
An undated image displaying Facebooks login page. — Unsplash
An undated image displaying Facebook's login page. — Unsplash

Meta has announced that it will start labelling the images generated by artificial intelligence tools from companies like OpenAI, Google, Midjourney, Microsoft, Shutterstock and Adobe.

Nick Clegg, Meta's president of global affairs, said the company in the coming months will deploy a technology across Facebook, Instagram and Threads that will detect and label such images, informing users that the images are digital creations. Meanwhile, the company already labels content generated by its own AI tools.

Once the new system is up and running, Meta will do the same for images created on services run by OpenAI, Microsoft (MSFT.O), opens new tab, Adobe , Midjourney, Shutterstock and Alphabet's Google, Clegg said.

Read more: Microsoft to offer Linux-like Sudo command on Windows server

The announcement offers an early preview of a developing set of standards that technology companies are creating to minimise the potential negative effects of generative AI technologies, which can produce deceptive yet convincing content with minimal input, Reuters reported.

This approach is based on a framework that has been established by some companies over the last ten years to collaborate on removing prohibited content from various platforms, such as images of mass violence and child exploitation.

Clegg expressed assurance in the company's choice to identify AI-generated content, noting that the tools for labeling audio and video content are more complex and are still in the development process.

"Even though the technology is not yet fully mature, particularly when it comes to audio and video, the hope is that we can create a sense of momentum and incentive for the rest of the industry to follow," Clegg said.

While currently there is no viable mechanism to label written text generated by AI tools like ChatGPT, he maintained that Meta, for the time being, would start requiring people to label their own altered audio and video content and would incur penalties over noncompliance.