Google has launched a brand new know-how that can embed watermarks in AI-generated photos. The watermarks will assist to establish photos which were created by AI, they usually can be used to trace the unfold of misinformation.
The watermarks are invisible to the bare eye, however they are often detected by software program. The watermarks embrace details about the AI mannequin that was used to create the picture, in addition to the date and time the picture was created.
Google says that the watermarks will assist to forestall the unfold of misinformation and deepfakes. Deepfakes are movies or photos which were manipulated to make it appear to be somebody is saying or doing one thing they by no means mentioned or did.
“We consider that watermarking AI-generated photos is a crucial step in stopping the unfold of misinformation and deepfakes,” mentioned a Google spokesperson. “We’re dedicated to working with the trade to develop and deploy this know-how in a approach that’s efficient and privacy-preserving.”
The usage of watermarks for AI-generated photos continues to be in its early phases, however it’s a promising know-how that has the potential to assist to forestall the unfold of misinformation.
Listed below are a number of the advantages of utilizing watermarks for AI-generated photos:
- They might help to establish photos which were created by AI.
- They might help to trace the unfold of misinformation.
- They might help to forestall the usage of AI-generated photos for malicious functions.
If you’re involved concerning the unfold of misinformation or deepfakes, you should utilize watermarks to assist to establish and observe these photos. You too can use watermarks to assist to forestall the usage of AI-generated photos for malicious functions.
Google shouldn’t be the one firm that’s engaged on watermarking AI-generated photos. Microsoft has additionally developed a watermarking know-how, and there are a variety of different corporations which might be engaged on related tasks.
The usage of watermarks for AI-generated photos continues to be in its early phases, however it’s a promising know-how that has the potential to assist to forestall the unfold of misinformation.