Social media service Snap said on Tuesday that it plans so as to add watermarks to AI-generated photos on its platform.
The watermark is a translucent model of the Snap brand with a sparkle emoji, and will probably be added to any AI-generated picture exported from the app or saved to the digicam roll.
The watermark, which is Snap’s brand with a sparkle, denotes AI-generated photos created utilizing Snap’s instruments. Picture Credit: Snap
On its help web page, the corporate stated eradicating the watermark from photos will violate its phrases of use. It’s unclear how Snap will detect the elimination of those watermarks. We now have requested the corporate for extra particulars and can replace this story after we hear again.
Different tech giants like Microsoft, Meta and Google have additionally taken steps to label or establish photos created with AI-powered instruments.
At the moment, Snap permits paying subscribers to create or edit AI-generated photos utilizing Snap AI. Its selfie-focused function, Goals, additionally lets customers use AI to boost their photos.
In a weblog publish outlining its security and transparency practices round AI, the corporate defined that it reveals AI-powered options, like Lenses, with a visible marker that resembles a sparkle emoji.
Snap lists indicators for options powered by generative AI. Picture credit: Snap
The corporate stated it has additionally added context playing cards to AI-generated photos created with instruments like Dream to higher inform customers.
In February, Snap partnered with HackerOne to undertake a bug bounty program geared toward stress-testing its AI image-generation instruments.
“We wish Snapchatters from all walks of life to have equitable entry and expectations when utilizing all options inside our app, significantly our AI-powered experiences. With this in thoughts, we’re implementing further testing to reduce doubtlessly biased AI outcomes,” the corporate stated on the time.
Snapchat’s efforts to enhance AI security and moderation come after its “My AI” chatbot spurred some controversy upon launch in March 2023, when some customers managed to get the bot to reply and discuss intercourse, ingesting and different doubtlessly unsafe topics. Later, the corporate rolled out controls within the Household Middle for folks and guardians to observe and prohibit their kids’s interactions with AI.
