WhatsApp has begun testing a groundbreaking feature that allows the creation of AI-generated stickers. This innovative feature was first noticed by participants in the Android beta program, according to WABetaInfo.
The new feature uses artificial intelligence to generate stickers based on text descriptions provided by users. This is a significant step forward in personalizing communication within the app, as it allows users to create unique and simple images that can be shared as stickers in individual chats or group conversations.
However, the specific generative AI model being used by WhatsApp for this feature remains unclear. Android Police has pointed out this uncertainty, noting that the functionality seems to operate similarly to models like Midjoureny or OpenAI’s DALL-E, which are known for generating images solely from text descriptions.
Despite the excitement surrounding this innovation, there are concerns about the potential misuse of such a feature. Given the ease with which images and stickers could be generated and distributed, there are fears about the propagation of misleading content. This concern is particularly relevant considering WhatsApp’s history of having to impose restrictions on the forwarding of viral messages during sensitive events like elections and mob incidents.
WABetaInfo also notes that the stickers produced through this AI-driven feature are “easily recognizable.” This suggests the possibility of a watermark or sign indicating their origin from generative AI models, similar to the approach adopted by Microsoft with OpenAI’s DALL-E model for Bing Chat.
As WhatsApp continues to test this novel feature, users eagerly await the opportunity to personalize their chats with AI-generated stickers. However, the company will need to ensure that adequate measures are in place to prevent the misuse of this feature.