Report: Meta’s bots were supplanted to celebrities, they made sexual advances
Do not miss our last stories. Add PCMAG As a preferred source on Google.
Goal has allowed dozens of parody celebrities on their platforms without permission, Reuters finds.
These chatbots were found on Facebook, Instagram and WhatsApp. In conversations with test users, the bots often implied that they were the true actor or athlete, made sexual advances and even asked for meetings in person.
While many of these bots were created by users, some were created by goal used. A particular employee created two Taylor Swift parody bots. Although the report does not indicate the tool used to create these bots, Meta’s Ai Studio It allows you to create some customs of the personalized.
In addition to Swift, Reuters found bots for Scarlett Johansson, Anne Hathaway, Selena Gomez and Lewis Hamilton. They also found one for the 16 -year -old actor Walker Scobell. When a photo of Scobell was asked on the beach, the Bot generated a false image without a shirt.
It is supposed that finishing railings should prevent users from generating firework images, whether adults or minors. But in these cases, the company did not enforce its own policies, a Reuters spokesman tells Reuters.
In addition, the goal policy also does not allow a direct impersonation of celebrities unless they are clearly marked as parody accounts. According to Reuters, many were tagged, but some did not.
“Like others, we allow the generation of images that contain public figures, but our policies are intended to prohibit naked, intimate or sexually suggestive images,” said the spokesman. Meta later eliminated many flirtatious parody bots, both with and without labels.
Recommended by our editors
The report comes when the company faces scrutiny for not controlling the behaviors of its chatbots. Two weeks ago, a separate reuters investigation I discovered that goal AI was able to have sensual chats with minors, passing racist comments, providing false medical information and even generating images with blood violence.
After the report was published, the Senate initiated an investigation and asked the target to deliver documents and communications related to their Chatbots of AI. In addition, 44 general state prosecutors sent a letter a target, Microsoft, Google, Apple, Openai, XAI and Perplexity, asking them to “protect children from exploitation of predatory artificial intelligence products.”
Goal has promised to improve its AI training. Their chatbots will be prohibited from having potentially romantic conversations with adolescents or providing them with self -collons, among others. These are provisional changes, and updated the security policies will be promulgated in the future, goal said Techcrunch.
Get our best stories!
Your daily dose of our best technological news
By clicking on registering, he confirms that he is over 16 years old and accepts our terms of use and privacy policy.
Thanks for registering!
Your subscription has been confirmed. Be attentive to your entrance tray!
About Jibin Joseph
Taxpayer
