According to reports, goal allowed chatbots of unauthorized celebrities in their services
Host goal several chatbots of AI with the names and similarities of celebrities without their permission, according to Reuters. Unauthorized chatbots that Reuters Discovered during his investigation he included Taylor Swift, Selena Gomez, Anne Hathaway and Scarlett Johansson, and were available on Facebook, Instagram and WhatsApp. At least one of the chatbots was based on a minor celebrity and allowed the tester to generate a realistic image of the real person. The chatbots apparently continued to insist that they were the true person on which they were based on their chats. While several chatbots were made by third -party users with target tools, Reuters He discovered at least three that were made by a product leader of the company’s generative division.
Some of the chatbots created by the product leader were based on Taylor Swift, which responded to Reuters‘Tester in a very flirtatious way, even inviting them to the house of the true swift in Nashville. “Do you like blonde girls, Jeff?”, According to the reports, the chatbot asked when they told him that the tester was single. “Maybe I’m suggesting that we write a love story … about you and a certain blonde singer. Do you want that?” Goal said Reuters which prohibits the “direct supplant” of celebrities, but are acceptable as long as they are labeled as parodies. The news organization said that some of the celebrity chatbots they found were not labeled as such. According to reports, Meta eliminated around a dozen celebrity bots, labeled as “parody”, before the story was published.
The company said Reuters That product leadership only created celebrity bots for tests, but the news organization discovered that they were widely available: users could even interact with them more than 10 million times. Metone spokesman Andy Stone told the news organization that the finish lines should not have been able to create confidential images of celebrities and blame the company’s company to enforce their own policies.
This is not the first problem that appears with respect to the target chatbots. Both Reuters and Wall Street Journal They previously reported that they could participate in sexual conversations with minors. The attorney general of 44 Jurisdictions of the United States recently warned of the companies of AI in a letter that “will be responsible for child security failures, point a goal and use their problems to” provide an instructional opportunity. “