The social media startup Snap said on Tuesday that it would be watermarking AI-generated photos on its network.

To identify an artificial intelligence (AI)-generated image, the platform is adding a little ghost logo with a glitter emblem. According to the corporation, the watermark will show up when the photo is saved to the camera roll or exported.

The company stated that erasing Snap’s Ghost watermark with sparkles is against its agreements on its support page. It’s unclear how Snap intends to identify the removal of the watermark. We’ve contacted the business to get further information, and we’ll update this article as soon as we do.

A number of other major internet companies, including Google, Microsoft, and Meta, have also taken action to identify or categorise photos produced using AI-powered techniques.

Currently, Snap offers a paid product called Dreams that focuses on selfies, as well as Snap AI, which enables users to create or modify AI-generated photographs.

The business described how it demonstrates generation AI-powered features like Lenses with visual cues like a glittering logo in a blog post detailing its safety and openness procedures surrounding AI.

In order to better enlighten the customer, the business also integrated context cards with AI-generated photos from technologies like Dream selfies.

In order to highlight its AI image-generation capabilities, Snap and HackerOne collaborated in February and modified a bug bounty programme. According to the business, while developing AI-powered glasses, a review mechanism has also been established to eliminate any issues.

“We want users of Snapchat from all backgrounds to have equal access to and expectations from all of our app’s features, especially our AI-powered experiences. In light of this, we’re conducting more testing to reduce the possibility of biassed AI outcomes,” the business stated on its blog.

Shortly after launching the “My AI” chatbot last year, Snapchat found itself in hot water. According to a Washington Post article, the bot was coming back.

unsuitable answers to users. Afterwards, the business released tools in the Family Centre that let parents and guardians keep an eye on and limit their adolescent’s use of AI.

Please be kind and stick to the topic of the conversation. You might see comments from our Community Managers, who will be identified by a “TechCrunch Staff” or “Staff” label, in an effort to promote civil and fruitful dialogues. After an article has been up for three days, we close comments to ensure the best possible user experience. Comments are always welcome on TechCrunch Finance’s Conversations chat boards. Check out our community guidelines for further details.

Leave a Reply

Your email address will not be published. Required fields are marked *