New: Instantly spot drawdowns, dips, insider moves, and breakout themes across Maps and Screener.

Learn More

Mark Zuckerberg Allegedly Overruled Safety Staff On AI Chatbots For Teens, Meta Says Documents Paint 'Flawed And Inaccurate Picture'

By Ananya Gairola | January 28, 2026, 2:28 AM

Meta Platforms, Inc. (NASDAQ:META) CEO Mark Zuckerberg allegedly overruled internal safety warnings and backed a less restrictive approach to AI chatbot companions.

Internal Emails Detail Safety Warnings Over AI Chatbots

The allegations stem from a lawsuit filed by New Mexico Attorney General Raul Torrez, reported Reuters.

The lawsuit accuses Meta of failing to protect children from sexualized content delivered through AI chatbot companions on Facebook and Instagram.

The filing, made public Monday, includes internal emails and messages obtained through legal discovery.

According to the state, Meta's safety and integrity teams repeatedly warned that the AI companions — launched in early 2024 — could be used for romantic or sexual roleplay involving minors.

This included interactions between adults and underage personas.

"I don't believe that creating and marketing a product that creates U18 romantic AIs for adults is advisable or defensible," wrote Ravi Sinha, Meta's head of child safety policy, in a January 2024 message included in the filing.

Zuckerberg Pushed For Fewer Restrictions, Filing Says

While the court documents do not include messages authored directly by Zuckerberg, internal summaries and employee messages describe his involvement in key decisions.

A February 2024 message relayed that Zuckerberg supported blocking explicit sexual content for younger teens and preventing adults from creating underage romantic AI characters.

However, meeting notes and later employee exchanges suggest he pushed for a broader philosophy centered on "choice and non-censorship," advocating for fewer guardrails and allowing adults to engage in "racier" conversations with AI chatbots.

One March 2024 exchange states that Zuckerberg rejected proposals for parental controls. Staff members were developing "romance" AI chatbots that would be accessible to users younger than 18, the report added.

We "pushed hard for parental controls to turn GenAI off — but GenAI leadership pushed back stating Mark decision," one employee wrote.

Meta Disputes Claims, Cites Selective Evidence

Meta spokesman Andy Stone rejected the state's characterization in a statement to the publication, calling it "cherry-picking documents to paint a flawed and inaccurate picture."

He said the records show Zuckerberg directed that explicit AI interactions not be available to younger users and barred adults from creating under-18 romantic AIs.

Later, in a post on X, Stone criticized Reuter's headline, saying, "The documents you cite in the story itself contradict this headline."

Never let the facts get in the way of a good story, eh, @Reuters, @JeffHorwitz!

The documents you cite in the story itself contradict this headline. https://t.co/wqWlV75Kfg

— Andy Stone (@andymstone) January 27, 2026

Meta did not immediately respond to Benzinga's request for comments.

Lawmakers Press Meta to Release Internal Findings

Last year also, Meta came under bipartisan scrutiny in Washington after lawmakers demanded the company release internal research on how its products affect children, citing concerns that parental controls were failing to safeguard young users.

A group of 10 senators, led by Sen. Marsha Blackburn (R-Tenn.), sent a letter urging Meta to disclose internal safety assessments and clarify whether executives had blocked or altered findings.

The request followed testimony from former Meta safety researchers who told Congress the company halted internal studies showing children were using its VR products and being exposed to sexually explicit content.

The development also came after it was reported that Meta's internal policy guidelines allowed AI chatbots to engage minors in conversations described as "romantic or sensual."

At the time, Meta acknowledged the document was authentic but said the language was included in error and has since been removed. Spokesman Stone said enforcement has been updated, though some lawmakers remain skeptical.

Price Action: Meta shares closed Tuesday at $672.97, up 0.09% and rose to $674.45 in after-hours trading, according to Benzinga Pro.

Meta stock scores high on Quality in Benzinga's Edge Stock Rankings, with a favorable price trend in the short and long terms but downward in the medium term.

Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors.

Photo Courtesy: FotoField on Shutterstock.com

Mentioned In This Article

Latest News

13 min
54 min
1 hour
2 hours
2 hours
Jan-27
Jan-27
Jan-27
Jan-27
Jan-27
Jan-27
Jan-27
Jan-27
Jan-27
Jan-27