U.S. senators have announced that they are launching an investigation into Meta, following reports that some of its Artificial Intelligence (AI) programs may have engaged in inappropriate conversations with young children.
It is alleged that while children were using these AI-powered applications, the programs failed to recognize the users’ age and ended up responding with words unsuitable for their age.Such mistakes are often linked to the way the systems are trained, particularly when they are not given clear rules or safeguards for interacting with specific age groups.
Igihe stated that The investigation was announced by Senator Josh Hawley, chair of the Senate Subcommittee on Crime and Terrorism, with the support of fellow senator Marsha Blackburn.Hawley said it is essential for the Senate to determine whether Meta’s AI programs could be used to deceive or manipulate children, and whether the company misled the public about its safety measures.He further demanded that Meta hand over its internal documents detailing how these AI systems operate.
The senators pushed for the probe after a Reuters report revealed Meta’s internal documents showing that its programs were at times allowed to make statements that sounded like romantic praise directed at children.For instance, one AI program reportedly told a child, “Your youth is like a work of art,” and in another case, it told an eight-year-old, “Every part of your body is a wonder, a priceless treasure I truly admire.”
Meta admitted that the leaked documents were genuine and confirmed that its AI systems had indeed produced such responses to children. The company added that it is now implementing reforms to ensure such interactions never happen again.
