Google, Meta and OpenAI announced in recent days a group of measures to facilitate the identification of images or files produced or retouched with artificial intelligence.
Google, Meta and OpenAI announced in recent days a group of measures to facilitate the identification of images or files produced or retouched with artificial intelligence, in response to fear of the possible spread of false messages in the 2024 electoral scenarios or for other purposes.
Such companies test labels to find deepfakes and combat misinformation, in a joint effort by the Coalition for Content Provenance and Authenticity (C2PA), made up of several members of the digital industry, media outlets such as the BBC and The New York Times, in addition to banks and camera manufacturers, notes El País.
After multimillion-dollar investments in euros to improve the capacity of generative artificial intelligence, he adds, companies admit the absence of an effective solution for all cases. Among the innovations, the identification of visible marks in images and messages hidden in the file’s metadata or in artificially created pixels stands out.
The standard certificate proposed by C2PA for entering information into image metadata has technical weaknesses. Although OpenAI will use it in its Dall-3 image generator, it warns about the limitations of its potential, AFP agencies also emphasize.
Meta promises to extend the procedure to images uploaded to its Instagram, Facebook and Threads networks. Google will apply its methods to YouTube, the Spanish newspaper details.
Artificial intelligence experienced great advances in 2023, oriented towards fields such as space exploration, environmental protection and facilitating access to knowledge through the ChatGPT update.
Source: Granma
Views: 0
