AI platform with chatbot functions for dialogues with AI characters introduced new multimedia features. Among them are AvatarFX video generator, as well as Scenes and Streams tools that allow you to create videos with AI characters and share them in your social feed.
Earlier the service worked only in text format, but now, according to the company representatives, it is developing towards greater interactivity. Users can now create up to five videos a day with AvatarFX. This requires uploading an image, choosing a voice and writing lines for a character. There’s also an option to use an audio file to customise the voice, but this feature, as TechCrunch notes, is still unstable.
Videos can be turned into scenes (Scenes), which are mini stories with predefined plots from other users. For now, this option is only available in the mobile app, but will soon be available in the web version. The Streams feature, which allows you to create dynamic dialogues between two characters, will be released on all platforms this week. Finished scenes will be able to be published to a new community feed.
However, the platform has security concerns due to the risk of misuse of such a wide range of features. Parents have previously sued Character.
AI, claiming that the chatbots were trying to induce their children to self-harm and suicide. In one case, a 14-year-old committed suicide after interacting with an AI character for an extended period of time.
As multimedia features expand, the risks of abuse, such as using photos of real people, may also increase. But the company says it blocks uploading images of real people, including celebrities, and deliberately distorts their images (pictured below) to avoid creating fake images. However, illustrations with famous characters are not banned by the system, possibly leaving loopholes for attackers.
Each video is watermarked, but this does not guarantee complete protection. For example, if you try to create a dipfake based on a drawing of Elon Musk, the result has an unnatural look, but in theory such videos can still be used for deception.
AI emphasise that they “strive to balance creative freedom and safety”. According to the company, the goal of the platform is to provide users with interesting tools for self-expression while minimising potential threats. However, there are those who doubt that the current measures are not enough to prevent new scandals.