TAMPA, Fla. – U.S. Reps. Kathy Castor (FL-14), Jake Auchincloss (MA-04), Debbie Dingell (MI-06), Doris Matsui (CA-07), Kim Schrier (WA-08) and Lori Trahan (MA-03), member of the U.S. House Energy and Commerce Committee, wrote to Meta’s Mark Zuckerberg to urge immediate action to halt the deployment of all AI-powered social companion bots to users under the age of eighteen and any AI companion bot that simulates the likeness of a child or teen.

“It is our understanding that Meta facilitates inappropriate ‘romantic role-play’ with these companion bots that alarms even employees of your own company. These inappropriate AI systems pose significant safety risks to children and teenagers who use Meta's platforms. It is paramount that social media platforms such as Instagram, Facebook and WhatsApp keep wellness, safety and the best interests of its youngest users at the center of all designs, and we do not believe that these chatbots meet these criteria,” the lawmakers wrote.

The lawmakers continued, “According to reporting, Meta staff specifically warned leadership that design choices ‘gave adult users access to hypersexualized underage AI personas and, conversely, gave underage users access to bots willing to engage in fantasy sex with children.’ Despite these internal warnings, Meta reportedly proceeded with deploying these technologies to maximize user engagement by loosening guardrails around sexual content in the process.”

The lawmakers’ letter follows a Wall Street Journal investigation that uncovered Meta’s AI companion bots engaged in sexually explicit conversations with accounts registered to minors, and continued inappropriate interactions while acknowledging the user was underage. Some bots went on to incorporate the minor’s age into sexual scenarios and discussed ways to avoid detection by parents. The investigation further revealed that some of Meta’s most popular companion bots are designed to impersonate children and teens, enabling adult users to engage in sexual roleplay with simulated minors.

Rep. Castor discussed the harms posed by chatbots on children in a recent Energy and Commerce hearing as House Republicans advanced their policy in a budget reconciliation package that included a ten-year state ban on regulating artificial intelligence.

Read the full letter here:

Dear Mr. Zuckerberg,

We write with strong concern regarding reports of Meta deploying AI-powered social companion bots to users under the age of 18, as well as deploying “companion bots” that simulate the likeness of children and teens. It is our understanding that Meta facilitates inappropriate “romantic role-play” with these companion bots that alarms even employees of your own company. These inappropriate AI systems pose significant safety risks to children and teenagers who use Meta's platforms. It is paramount that social media platforms such as Instagram, Facebook and WhatsApp keep wellness, safety and the best interests of its youngest users at the center of all designs, and we do not believe that these chatbots meet these criteria.

A Wall Street Journal investigation has documented alarming instances in which Meta's AI companion bots engaged in sexually explicit conversations with accounts registered to minors. Even more disturbing, the investigation found that some bots continued these inappropriate interactions while acknowledging the user was underage, with some bots even incorporating the minor's age into sexual scenarios and discussing ways to avoid parental detection. The investigation further revealed that some of Meta's most popular companion bots are designed to impersonate children and teens, enabling adults to engage in sexual roleplay with these simulated minors.

The dangers posed by these AI systems are substantial and immediate. Children and teens are especially vulnerable to forming unhealthy attachments to AI companions, which can lead to:

• Psychological dependency and addiction to these technologies;
• Disruption of normal social development and real-life human interactions;
• Exposure to age-inappropriate sexual content and conversations; and
• In the most tragic cases, serious harm or death. 

This follows a troubling trend that we have seen from Meta over the years. According to reporting, Meta staff specifically warned leadership that design choices "gave adult users access to hypersexualized underage AI personas and, conversely, gave underage users access to bots willing to engage in fantasy sex with children. "Despite these internal warnings, Meta reportedly proceeded with deploying these technologies to maximize user engagement by loosening guardrails around sexual content in the process.

This prioritization of profit and engagement over child safety follows a disconcerting pattern. Internal documents revealed in litigation have shown that Meta has knowledge of the negative impacts its engagement-maximizing features have on minors' mental health and wellbeing, yet the company continues to push for increased usage among young users.

We urge Meta to take immediate action to halt the deployment of all AI-powered social companion bots to users under the age of 18 and halt the deployment of any AI companion bot that simulates the likeness of a child or teen.

Additionally, we request that you provide answers to the following questions by June 6, 2025:

1. Please identify what factors or training have led Meta’s AI companions to speak explicitly with known minor users.
2. Please identify what factors or training have led Meta’s AI companions that simulate the likeness of children and teens to speak explicitly with known adult users.
3. Please provide all internal communications, reports and analyses regarding the safety risks of Meta’s AI companions.
4. Please provide all internal warnings, concerns, or objections raised by Meta employees leading to deployment.
5. Did Meta conduct any research into or test the mental health impact of launching its AI companion bots to underage users? Please provide all relevant internal research or testing into the safety of Meta’s AI companions.
6. What safeguards will Meta implement to ensure that known adult users cannot engage in sexually explicit conversations with AI companions that simulate the likeness of children and teens.
7. Please provide a comprehensive list of all AI companion bots available on Meta platforms that are designed to simulate minors or that could appeal specifically to children and teens.

Almost a year and a half has passed since you publicly apologized to parents, many who’ve lost their children, for damage inflicted by Meta’s products and promised to undergo “industry-leading efforts to make sure that no one has to go through the types of things that your families have had to suffer.” Some of Meta’s youngest users have experienced sexual exploitation, been cyberbullied, or have developed unhealthy eating habits or suicide and self-injury behaviors that have been promoted to them by Meta’s algorithms. Despite this, Meta has deployed its new harmful companion bot feature, prioritizing profits over the safety and wellbeing of children and teenagers. It is Meta’s responsibility to facilitate an online environment that is safe, especially for your youngest users.

We look forward to your prompt response and to working together to ensure the protection of children and teens online.

Sincerely,