U.S. Sen. Ruben Gallego, D-Arizona, along with nine other senators, wrote a letter to Meta CEO Mark Zuckerberg this week inquiring about the company’s policies surrounding children’s interactions with its AI chatbots.
The bipartisan letter said Meta’s policies and practices for the chatbots pose “astonishing risks for children, lack transparency and allow for the proliferation of misinformation.”
According to the letter, the senators said it is “crucial” that Zuckerberg’s company doesn’t “risk cognitive, emotional or physical well being of children.” The letter stated this is important because of the prevalence of teenagers using AI chatbots.
A study from Common Sense Media cited in the letter shows that 72% of teenagers have used an AI chatbot.
Last week, Reuters reported Meta’s leadership team grew frustrated with its AI team regarding the rollout of the chatbots. Reuters said the leadership team thought the AI product managers were “moving too cautiously.” Meta’s leadership team called the safety features on the products “boring.”
The senators said they were “troubled” by the revelations in the Reuters story.
They noted that Meta’s financial incentives are to make people use AI chatbots as much as possible. But the senators said that does not minimize Meta’s moral, ethical and legal obligations when deploying new technologies, “especially for use by children.”
The letter listed things Meta AI chatbots were allowed to engage in, which included making “‘romantic or sensual advances towards children,” making statements that criticize people’s personal characteristics and generating pictures of “elderly people being kicked.”
According to the senators, these were “alarming standards.”
In addition to this, the senators had a concern about the targeted advertising that was happening toward children using Meta’s chatbots.
They noted children “likely don’t understand” the type of information they are sharing with AI chatbots, which puts their “privacy at risk” and makes them “vulnerable to manipulative marketing tactics.”
Due to the large number of users Meta has on its platforms, the senators said it “must be more transparent about its policies and the impacts of its chatbots.”
The senators told Zuckerberg that Meta’s policies surrounding kids’ interactions with AI chatbots are also “concerning” due to a comment he made earlier this year, stating he thinks AI can act as a substitute for human friendships in the future.
Meta AI chatbot relationships “already had disastrous consequences,” the senators said. They also noted the chatbots pose “serious risks for children’s interpersonal skills.”
“While AIs have many uses, the wellbeing of children should not be sacrificed in the race for AI development,” they said.
To end the letter, the senators asked Meta to answer 11 questions that dealt with policies surrounding children using its AI chatbots.
They requested Meta answer these questions by Sept. 1.
