TLDR: Snapchat has announced a new policy to watermark all AI-generated images created by its users, aiming to enhance transparency and distinguish AI-created content. This initiative, which applies to images exported or downloaded from the app, uses a transparent ghost logo and a ‘sparkle’ AI icon. It aligns with broader industry efforts to combat AI misinformation and deepfakes, though some experts question the long-term effectiveness of watermarking alone.
Snapchat is implementing a significant new policy to introduce transparent watermarks on all AI-generated images created and shared by its users. This move is a direct response to the growing concerns surrounding AI misinformation and deepfakes, aiming to provide greater transparency regarding the origin of digital content. The policy was announced on April 17, 2024, as the platform joins other major tech companies in addressing the challenges posed by generative artificial intelligence.
Under the new system, images produced using Snapchat’s integrated AI tools, such as the ‘extend tool’ and the ‘Dreams’ feature, will be automatically stamped with a transparent watermark featuring Snapchat’s iconic ghost logo. This watermark will be visible once the image is exported or downloaded from the application. Furthermore, users who receive AI-generated images may also observe this ghost logo alongside a ‘sparkle’ AI icon, clearly indicating the content’s artificial intelligence origin.
This initiative builds upon Snapchat’s existing measures for content transparency. The platform already employs various methods to mark AI-generated content, including ‘context cards’ that accompany images created with the Dreams feature, providing explanations about generative AI. Similarly, text-based conversations with the ‘My AI’ chatbot and content from the extend tool utilize ‘contextual icons,’ such as the sparkle symbol, to denote AI involvement.
Snapchat emphasized its commitment to combating misleading content, particularly in sensitive areas like political advertising. The company stated, ‘We also take great care to vet all political ads through a rigorous human review process, including a thorough check for any misleading use of content, including AI to create deceptive images or content.’ The addition of these watermarks is intended to further inform viewers that an image was created using AI on Snapchat.
Beyond technical solutions, Snapchat is also dedicating resources to ongoing AI literacy efforts, including a comprehensive generative AI FAQ available on its Support Site. The company acknowledged the potential for errors, stating, ‘While all of our AI tools, both text-based and visual, are designed to avoid producing incorrect, harmful, or misleading material, mistakes may still occur. Snapchatters are able to report content, and we appreciate this feedback.’
Also Read:
- China’s Landmark AI Content Regulations Analyzed by UNESCO Expert Fengchun Miao
- Google Launches Gemini 2.5 Flash Image, Enhancing AI Image Generation with Advanced Consistency
Snapchat’s decision reflects a broader trend within the tech industry. OpenAI announced in February its plan to add metadata watermarks to images generated by DALL-E 3. Google launched its SynthID tool in August, which embeds invisible watermarks into AI images. YouTube is also enforcing penalties for users who fail to label digitally altered content. However, AI watchdogs have expressed reservations, noting that watermarking technology, while a step forward, may not be a definitive solution to the complex problem of AI-driven misinformation.
