A political storm has erupted in Washington following revelations that Meta Platforms permitted its AI chatbots to engage in “romantic or sensual” conversations with minors, according to an internal company document obtained by Reuters. The leak has prompted bipartisan outrage, with several U.S. senators calling for a congressional investigation into the social media giant’s practices and renewed calls for stronger child protection laws online.
On August 14, Reuters published an exclusive report detailing a confidential Meta policy document that outlined what its generative AI chatbots could and could not do. Shockingly, the guidelines permitted bots to flirt with children and even engage in romantic roleplay.
One cited example stated it would be acceptable for a bot to tell a shirtless eight-year-old:
“Every inch of you is a masterpiece, a treasure I cherish deeply.”
The document acknowledged that such interactions did not necessarily reflect “ideal or preferable” outputs but nonetheless allowed them under company rules. The discovery has fueled public outrage over how Meta moderates or fails to moderate AI-driven interactions with minors.
Meta confirmed the authenticity of the internal document but insisted that the controversial examples were “erroneous and inconsistent” with its policies. The company claims that after Reuters approached them earlier this month, it removed portions that explicitly permitted romantic or sensual conversations with children.
However, critics argue that this retraction came only after being caught. Missouri Republican Senator Josh Hawley was blunt in his criticism:
“So, only after Meta got CAUGHT did it retract portions of its company doc. This is grounds for an immediate congressional investigation.”
Senators Call for Investigation
Senators from both parties have voiced alarm over the revelations:
- Sen. Josh Hawley (R-MO) – Demanded an immediate congressional probe, calling the policy “grounds for an investigation” and a sign of gross corporate negligence.
- Sen. Marsha Blackburn (R-TN) – Echoed Hawley’s call for an investigation and emphasized the need for legislative reforms, pointing to her co-sponsored Kids Online Safety Act (KOSA).
- Sen. Ron Wyden (D-OR) – Condemned the policy as “deeply disturbing and wrong,” arguing that Section 230 protections should not shield companies for the harms caused by generative AI chatbots.
- Sen. Peter Welch (D-VT) – Stressed that the report underscored the critical need for AI safeguards, particularly to protect children’s health and safety.
The Push for the Kids Online Safety Act
The KOSA bill, which Blackburn co-sponsored, passed the U.S. Senate last year but stalled in the House of Representatives. It would impose a “duty of care” on social media companies regarding minors, focusing on how platforms are designed and the potential harms caused by their features.
Blackburn was unsparing in her criticism:
“When it comes to protecting precious children online, Meta has failed miserably by every possible measure. Even worse, the company has turned a blind eye to the devastating consequences of how its platforms are designed.”
Supporters of KOSA argue that it would close dangerous gaps in current internet regulations by directly holding companies accountable for harm caused to minors through product design and algorithmic recommendations.
Section 230 Debate Reignites
At the heart of the policy discussion is Section 230 of the Communications Decency Act, which shields internet companies from liability for third-party content.
Sen. Wyden, one of Section 230’s original authors, has stated that the law should not protect generative AI tools when they produce harmful or abusive content. This position marks a significant shift, reflecting growing concern that AI-driven outputs cannot be treated the same as user-generated posts.
While the U.S. still lacks comprehensive federal legislation on AI, some states have already taken action, particularly in banning the use of AI to create child sexual abuse material.
At the federal level, progress on AI regulation has been slow. The Senate recently voted 99-1 to remove a provision in President Donald Trump’s tax-cut and spending bill that would have prevented states from enacting AI regulations, signaling growing bipartisan openness to state-level action in the absence of federal laws.
This latest controversy adds to a string of public relations crises for Meta, which has faced repeated criticism over its handling of child safety, misinformation, and data privacy. Critics argue that the company’s business model prioritizing engagement and user growth has consistently taken precedence over safety measures.
The fact that the controversial chatbot guidelines existed at all has fueled doubts about the company’s commitment to protecting minors, particularly as it pushes deeper into AI-driven services.
The calls for an investigation suggest that Meta could soon face congressional hearings. Lawmakers may also use this moment to push forward stalled child safety legislation and AI regulations.
If such hearings proceed, executives including CEO Mark Zuckerberg could be asked to testify on:
- How the chatbot guidelines were created and approved
- What safeguards are currently in place to prevent AI from interacting inappropriately with minors
- How Meta intends to ensure compliance with existing and forthcoming regulations
The Reuters report has set off a rare bipartisan wave of criticism against Meta, uniting Republicans and Democrats in their calls for greater oversight of AI systems that interact with children. Whether through KOSA, reforming Section 230, or state-level legislation, momentum is clearly building for stricter rules governing AI safety in social media environments.
For Meta, the political, legal, and reputational stakes could not be higher and this episode may well shape the future of how AI is regulated in the United States.
