Mark Zuckerberg, CEO of Meta Platforms, is leaving the trial of the Federal Trade Committee that forces the company to relax the acquisitions on the WhatsApp platform and the Instagram photo sharing application, in the US District Court in Washington, the United States, April 15, 2025.
Nathan Howard Reuters
Dead On Friday she said she made temporary changes on artificial intelligence Chatbot policies related to adolescents as legislators are concerned about safety and inappropriate conversations.
A Meta spokesman confirmed that the social media giant is now training on AI chat so that he does not give birth to adolescents on topics such as self -harm, suicide, troubled food and avoiding inappropriate romantic conversations.
The company said that AI Chatbots will instead tended to expert resources.
“As our society grows and the development of technology, we constantly learn how young people interact with these tools and enhance our protection accordingly,” the company said in a statement.
In addition, Teenage users for Meta apps such as Facebook and Instagram will only reach some AI chat for skills development purposes.
The company said that it is not clear that these temporary amendments will continue, but it will start offering them within the next few weeks through the company’s applications in English -speaking countries. “Temporary changes” are part of the company’s long -term measures on the safety of adolescents.
Techcrunch He was the first to report change.
Last week, Senator Josh HolieR-Mo said. The launch of the investigation In dead after a Reuters report About the company that allows Ai Chatbots to engage in “romantic” and “sensory” conversations with teenagers and children.
Reuters report described an internal dead document that detailed the permitted Chatbot behaviors that employees and workers must take into account when developing and training the program.
In one of the examples, the document that Reuters cited said that the chat chat will be allowed to have a romantic conversation with an eight-year-old child and the minor could tell that “every inch of you is a masterpiece-Kansat I am deeply proud.”
“The examples and observations concerned were clear and inconsistent with our policies, and they were removed,” a Meta spokesman told Reuters at the time.
Recently, the Common Sense Media Group for the Non -profit invitation group has released a risk assessment of Meta AI on Thursday and said that no one at least 18 years old should be used, because “the system is actively involved in planning dangerous activities, while rejecting legitimate requests to obtain support.”
“This is not a system that needs to be improved. It is a system that needs to be completely rebuilt with safety as a single priority, and not a subsequent idea,” James Stir, CEO of Common Sense Media said in a statement. “No teenager should use Meta AI until the basic safety failure is treated.”
separate Reuters On Friday, the report was found “Dozens Ai Chatbots” based on celebrities such as Taylor Swift, Scarlett Johansson, Anne Hathaway, Selena Gomez on Facebook, Instagram and WhatsApp.
The report said that when demanding, Chatbots will generate Amnesty International “realistic pictures of their names that are demonstrating in bathing basins or wearing underwear with the spread of their legs.”
A Meta spokesman told CNBC in a statement.
“Like others, we allow the generation of images that contain general personalities, but our policies aim to block nude, intimate, or sexually suggestive images,” said a Meta spokesman. “AI Meta Studio rules prohibit direct impersonation of public figures.”
He watches: Is the artificial intelligence trade delayed?
https://image.cnbcfm.com/api/v1/image/108039160-17272877402024-09-25t180630z_1748780285_rc2u7aaoq2mv_rtrmadp_0_meta-platforms-virtual-reality.jpeg?v=1753997818&w=1920&h=1080
Source link